Stay tuned...

`);\n});\n```\n\n## Architecture Validation\n\nBefore proceeding, verify your implementation follows the architecture:\n\n- Services only make direct API calls (no environment variables, no orchestration)\n- Controllers handle business logic and coordinate services\n- Routes only handle HTTP concerns and call controllers\n- No service calls directly from routes\n- Environment variables handled in controllers, not services\n- Consistent response patterns across all layers\n\n### Expected Behavior\n\n**Important**: for the health endpoint JSON below, get the project name from the `yaml` code block in `/_townie/00-setup.md`; it is the value of the `project` key in the `yaml` code.\n\nThe health endpoint should return:\n\n```\n{\n \"status\": \"ok\",\n \"timestamp\": \"2025-07-14T18:26:28.270Z\",\n \"val\": {{project}},\n \"databases\": {\n \"configured\": true,\n \"{{db1}}\": \"connected\",\n // other databases\n }\n}\n```\n\nThis architecture is intended to create clear separation of concerns and should make the codebase maintainable and testable.\n\n---\n\n**Townie, stop here! Before proceeding to additional steps, confirm that this step is working correctly.**\n\n**If there was an error regarding environment variables**: remind the user to upload `vars.env` to the Environment Variables section of this Val, and to then let you know when that step is complete so you can retry.\n\n**If everything is working as intended**: conclude this step with these messages:\n\n1. Briefly summarize the work you just did and why it's important\n1. Provide the URL to the health endpoint so the user can see that the endpoint is live. (Tell the user that they can copy and paste that URL into a new tab to see it live.)\n\n---\n\n### Next step: web route authentication\n\nTell the user that the next step is to add authentication to the web routes of this Val, and that they should copy this line and paste it into Townie:\n\n> Add auth to this Val with what you see in `/_townie/02-auth.md`\n",[],[154,183],{"_102":190,"_95":191,"_155":192,"_157":156,"_158":137,"_159":160,"_161":162,"_104":193,"_99":32,"_164":177,"_166":194,"_167":22,"_168":-5,"_169":195,"_98":196},"d0f9956a-7951-11f0-8810-0224a6c84d84","02-auth.md","0198aa63-0ee2-725e-8c9f-80d0e5e060b2",["D",1755205275298],"# Complete Authentication Implementation Instructions\n\n## Townie instructions (Townie should follow these)\n\nRead through all of these instructions in order to understand the architecture of this Val, and implementation details important to this Val, and then once you have it all in your memory you should get to work.\n\nIf anything is unclear, you should stop and ask me for clarification before proceeding. Do not write code if you are unsure of what is expected.\n\n**Important:** there are checkpoints in these instructions where you are directed to stop and ask permission before proceeding. Follow those instructions and do not proceed without permission.\n\n## LastLogin Built-in Features\n\n**Important:** LastLogin automatically handles certain routes and authentication flows:\n\n- **`/auth/logout`** - LastLogin intercepts this route and handles logout automatically\n- **No custom logout route needed** in your Hono app\n- **Simply link to `/auth/logout`** and LastLogin will clear the session and redirect\n- **Authentication headers** - LastLogin adds `X-LastLogin-Email` header for authenticated requests\n\n**Authentication Flow:**\n\n1. **Unauthenticated requests** → LastLogin checks for session\n2. **No session** → Returns 401 or redirects to login\n3. **Valid session** → Adds `X-LastLogin-Email` header and passes to your app\n4. **`/auth/logout` requests** → LastLogin clears session automatically (never reaches your app)\n\n## Step 1. Core Architecture Setup\n\n### Main Application (main.tsx)\n\n**Add global authentication middleware to `main.tsx`:**\n\n```typescript\nimport { lastlogin } from \"https://esm.town/v/stevekrouse/lastlogin_safe\";\n\n// Import route modules\nimport authCheck from \"./backend/routes/authCheck.ts\";\n// other route modules\n\nconst app = new Hono();\n\n// Apply global authentication middleware to ALL routes\napp.use(\"*\", authCheck);\n\n// other routes\n\n// Export with LastLogin wrapper\nasync function handler(request: Request): Promise {\n return app.fetch(request);\n}\n\n// Wrap and export the main HTTP handler with lastlogin\nexport default lastlogin(handler);\n```\n\n### Dashboard Route Organization\n\n**Best Practice: Separate the root dashboard route from main.tsx**\n\nInstead of defining the root route directly in `main.tsx`, create a dedicated dashboard route file:\n\n- **Create**: `/backend/routes/views/dashboard.tsx`\n- **Purpose**: Handle the authenticated user's main dashboard/landing page\n- `import dashboardRoute from \"./backend/routes/views/dashboard.tsx\";`\n- **Mount in main.tsx**: `app.get(\"/\", dashboardRoute)`\n- **Note**: Dashboard is placed in `/routes/views/` because it's a user-facing view, not an API endpoint\n\nThis approach:\n\n- Keeps `main.tsx` focused on app setup and middleware\n- Makes the dashboard logic easier to maintain and test\n- Follows the modular architecture pattern with views separated from API routes\n- Places user-facing interfaces in `/routes/views/` for better organization\n- Allows for easier dashboard enhancements without touching core routing\n\nThe dashboard route should:\n\n- Display user information (email from `c.get(\"userEmail\")`)\n- Display the `/api/health` endpoint results as raw JSON in a
 tag; e.g., JSON.stringify(data, null, 2)\n- Include a link to the raw `/api/health` endpoint below the JSON display for easy access and testing\n- Include logout functionality via `/auth/logout`\n- Show a simple and professional welcome interface with minimal styling\n\n### URL Utility for Absolute URLs\n\n**Create `/backend/utils/url.ts`:**\n\nBefore implementing the dashboard, create a utility function for constructing absolute URLs when needed (like for sharing URLs externally):\n\n```typescript\nexport function getAbsoluteUrl(c: any, path: string): string {\n  const protocol = c.req.header(\"x-forwarded-proto\") || \"https\";\n  const host = c.req.header(\"host\");\n  return `${protocol}://${host}${path}`;\n}\n```\n\n**Usage in dashboard and other routes:**\n\n- Import: `import { getAbsoluteUrl } from \"../../utils/url.ts\";`\n- Use for external sharing: `const dashboardUrl = getAbsoluteUrl(c, '/');`\n- Continue using relative URLs for internal links: `href=\"/api/health\"`\n\n**When to use absolute vs relative URLs:**\n\n- **Absolute URLs**: External sharing, webhooks, emails, API documentation\n- **Relative URLs**: Internal navigation, forms, API calls within the app\n\n### Create `/backend/routes/authCheck.ts`\n\n**Authentication Middleware:**\n\n- Retrieve the user's email: `const email = c.req.header(\"X-LastLogin-Email\")`\n- If email exists, the user is authenticated - store in context and continue to your app\n- If no email, show login page with LoginWithGoogleButton\n- **Note:** You don't need to handle `/auth/logout` - LastLogin does this automatically\n\n**Login Page Requirements:**\n\n- `import { LoginWithGoogleButton } from \"https://esm.town/v/stevekrouse/LoginWithGoogleButton\"`\n- Use the `` React Component, optionally supplying the `text` attribute\n- Add \"via LastLogin\" underneath LoginWithGoogleButton, centered, secondary text, linking to https://lastlogin.io/\n- Pass the email from the server to the client-side code if using React hydration or similar techniques\n\n**For authenticated users:**\n\n- Store email in context: `c.set('userEmail', email)`\n- In your authenticated views, add logout functionality: `Logout`\n\n**Public Routes Whitelist:**\n\n- Add a `PUBLIC_ROUTES` array at the top for routes that should be accessible without authentication\n- **Don't include `/auth/logout`** in the whitelist (LastLogin handles it before your middleware runs)\n- Use for truly public routes like: `/demo/:id`, `/api/health`, static assets, etc.\n\nIf you encounter problems loading the React button, try this simple approach for authCheck.ts:\n\n```typescript\n// Public routes that don't require authentication\nconst PUBLIC_ROUTES = [\n  // Add public routes here as needed\n  // '/demo/:id',  // Example: public demo viewing\n  // '/api/public', // Example: public API endpoints\n];\n\n// Helper function to check if a route is public\nfunction isPublicRoute(path: string): boolean {\n  return PUBLIC_ROUTES.some((route) => {\n    // Handle exact matches\n    if (route === path) return true;\n\n    // Handle wildcard patterns (basic implementation)\n    if (route.includes(\"*\")) {\n      const pattern = route.replace(\"*\", \".*\");\n      return new RegExp(`^${pattern}$`).test(path);\n    }\n\n    // Handle route parameters (basic implementation)\n    if (route.includes(\":\")) {\n      const pattern = route.replace(/:[^/]+/g, \"[^/]+\");\n      return new RegExp(`^${pattern}$`).test(path);\n    }\n\n    return false;\n  });\n}\n\nexport default async (c, next) => {\n  // Allow public routes to bypass authentication\n  if (isPublicRoute(c.req.path)) {\n    await next();\n    return;\n  }\n\n  const email = c.req.header(\"X-LastLogin-Email\");\n\n  if (!email) {\n    // Show login page if not authenticated\n    return c.html(`\n      \n      \n      \n        Your App - Login Required\n      \n      \n        \n          Sign in
\n          Sign in to access your app.
\n          \n            Loading login button...
\n          
\n          \n            via LastLogin\n          
\n        
\n\n        \n      \n      \n    `);\n  }\n\n  // If authenticated, store email in context for use by routes\n  c.set(\"userEmail\", email);\n\n  // Continue to the next middleware/route\n  await next();\n};\n```\n\n## Common Gotchas\n\n**Logout Route:**\n\n- ❌ Don't create a custom `/auth/logout` route in your Hono app\n- ✅ Just link to `/auth/logout` - LastLogin handles it automatically\n- The logout link will work for authenticated users without any custom code\n- LastLogin intercepts `/auth/logout` at the wrapper level before it reaches your middleware\n\n**Public Routes:**\n\n- Use the whitelist in authCheck.ts for routes that should be accessible without authentication\n- `/auth/logout` doesn't need to be whitelisted (LastLogin handles it before your middleware runs)\n- Examples of routes that should be whitelisted: public demos, health checks, static assets\n\n**Authentication Headers:**\n\n- Authenticated requests will have `X-LastLogin-Email` header\n- Unauthenticated requests will not have this header\n- LastLogin wrapper handles session validation before your code runs\n\n**Error Handling:**\n\n- If LastLogin can't validate a session, it returns 401 before reaching your app\n- Your authCheck middleware only runs for requests that pass LastLogin's validation\n- No need to handle session expiration - LastLogin manages this\n\n---\n\n**Townie, stop here! Before proceeding to additional steps, confirm that this step is working correctly.**\n\n**If everything is working as intended**: conclude this step with these messages:\n\n1. Briefly summarize the work that was just completed and why it's important\n1. Use the getAbsoluteUrl utility to construct the dashboard URL and provide it to the user so they can see that the dashboard is live. Tell the user that they can copy and paste that URL into a new tab to see it live.\n1. Tell the user that they can log into the dashboard using their Google email address.\n\n---\n\n### Next step: webhook authentication\n\nTell the user that the next step is to add webhook authentication to this Val. Explain what webhooks are why it's important to authenticate them.\n\nLastly, tell the user to copy this line and paste it into Townie:\n\n> Add webhook auth to this Val with what you see in `/_townie/03-webhooks.md`\n",[],[154,191],{"_102":198,"_95":199,"_155":200,"_157":156,"_158":137,"_159":160,"_161":162,"_104":201,"_99":32,"_164":177,"_166":202,"_167":22,"_168":-5,"_169":203,"_98":204},"d0f9a28a-7951-11f0-8810-0224a6c84d84","03-webhooks.md","0198aa63-0ee2-725e-8c9f-85a7e7bb5cea",["D",1755205275298],"# Webhook Authentication Architecture Instructions\n\n## Townie instructions (Townie should follow these)\n\n## Middleware Order\n\nPlace authentication middleware in the following order:\n\n1. **Notion webhook auth middleware** (X-API-KEY for `/tasks/*`)\n2. **Global user auth middleware** (for other protected routes)\n3. **Route handlers**\n\n## Webhook Authentication\n\n### Scope\n\n- Apply only to routes under /tasks/\\*\n- All /tasks endpoints expect POSTs from Notion webhooks\n\n### Implementation\n\n- Create `/backend/routes/webhookAuthCheck.ts`\n- Extract X-API-KEY header from incoming requests\n- Compare against NOTION_WEBHOOK_SECRET environment variable\n- Use constant-time comparison to prevent timing attacks\n- Return 401/403 for authentication failures with generic error messages\n- Log failed authentication attempts for monitoring\n\n### Route Implementation Requirements\n\n**CRITICAL:** All webhook routes MUST use POST method:\n\n```\n// ✅ CORRECT - Webhooks are POST requests\napp.post(\"/webhook-endpoint\", async (c) => {\n  // Handle webhook\n});\n\n// ❌ WRONG - This won't work with webhooks\napp.get(\"/webhook-endpoint\", async (c) => {\n  // Webhooks can't reach this\n});\n```\n\nWhy: Notion (and most webhook providers) send POST requests, never GET.\n\n### Middleware Application (EXACT CODE)\n\nApply middleware in this EXACT order in main.tsx:\n\n```\n// 1. Webhook auth - ONLY for /tasks/* routes\napp.use(\"/tasks/*\", webhookAuth);\n\n// 2. User auth - for ALL routes EXCEPT /tasks/*\napp.use(\"/api/*\", authCheck);\napp.use(\"/views/*\", authCheck);\napp.use(\"/\", authCheck);\n\n// Add root route - now serves the authenticated dashboard\napp.get(\"/\", dashboardRoute);\n\n// 3. Mount routes AFTER middleware\napp.route(\"/tasks\", taskRoutes);\n```\n\n**Important:** We've changed the global authentication from app.use(\"\\*\") to\nmore specific auth route handling.\n\n**Before:**\n\n```\napp.use(\"*\", authCheck);  // Applied to ALL routes including /tasks/*\n```\n\n**After:**\n\n```\napp.use(\"/api/*\", authCheck);   // Applied to API routes\napp.use(\"/views/*\", authCheck); // Applied to view routes\napp.use(\"/\", authCheck);        // Applied to root route\n```\n\n**This ensures:**\n\n✅ `/` (root) still requires user authentication ✅ `/api/*` routes still\nrequire user authentication ✅ `/views/*` routes still require user\nauthentication ✅ `/tasks/*` routes only use webhook authentication (no user\nauth)\n\n**Common mistake:** Using app.use(\"\\*\", webhookAuth) - this applies to all\nroutes incorrectly.\n\n### Flow Logic\n\n- Request to /tasks/\\* → Check X-API-KEY → If valid, proceed to route handler\n  (skip global auth)\n- Request to /tasks/\\* → Check X-API-KEY → If invalid, return 401 (never reaches\n  global auth)\n- Request to other routes → Skip webhook auth → Go to global auth → Then route\n  handler\n\n### Testing Webhook Authentication\n\nTest in this exact order:\n\n1. **Test without header** (should fail):\n\nPOST /tasks/test Expected: 401 \"Authentication required\"\n\n1. **Test with wrong key** (should fail):\n\nPOST /tasks/test Headers: {\"X-API-KEY\": \"wrong-key\"} Expected: 403\n\"Authentication failed\"\n\n1. Test with correct key (should succeed):\n\nPOST /tasks/test\\\nHeaders: {\"X-API-KEY\": \"your-configured-secret-value\"} Expected: 200 (reaches\nyour route handler)\n\n**If step 3 returns 404:** Your route is probably defined as GET instead of\nPOST.\n\n### Debugging Webhook Auth Issues\n\nAdd this temporary debug endpoint:\n\n```\napp.get(\"/debug-webhook\", (c) => {\n  const secret = Deno.env.get(\"NOTION_WEBHOOK_SECRET\");\n  return c.json({\n    hasSecret: !!secret,\n    secretLength: secret?.length || 0,\n  });\n});\n```\n\n**Expected response:** {\"hasSecret\":true,\"secretLength\":>0}\n\nAdd this catch all endpoint:\n\n```\n// Catch-all handler for undefined webhook endpoints\n// This must be the LAST route defined to catch unmatched paths\napp.post(\"*\", (c) => {\n  const path = c.req.path;\n  const method = c.req.method;\n  const returnObj = {\n    error: \"Endpoint not found\",\n    path: path,\n    method: method,\n    availableEndpoints: [\n      \"GET /tasks/debug-webhook\",\n      \"POST /tasks/test\",\n      \"POST /tasks/notion-webhook\",\n    ],\n  };\n  console.log(returnObj);\n\n  return c.json(returnObj, 404);\n});\n```\n\n### Complete Working Example\n\nHere's a complete working webhook endpoint:\n\n```\n// backend/routes/tasks/_tasks.routes.ts\nimport { Hono } from \"npm:hono@3.12.12\";\n\nconst app = new Hono();\n\napp.post(\"/notion-webhook\", async (c) => {\n  // At this point, webhook auth has already passed\n  const body = await c.req.json();\n  console.log(\"Notion webhook received:\", body);\n\n  // Process the webhook...\n\n  return c.json({ success: true });\n});\n\nexport default app;\n```\n\nTest with:\n\nPOST /tasks/notion-webhook Headers: {\"X-API-KEY\":\n\"your-configured-secret-value\"} Body: {\"test\": \"data\"} Expected: 200 {\"success\":\ntrue}\n\n### 4\\. Add Simplified Webhook Testing to Dashboard\n\nUpdate `/backend/routes/views/dashboard.tsx`:\n\n**Import the URL utility in the dashboard:**\n\nAdd this import at the top of `/backend/routes/views/dashboard.tsx`:\n\n```typescript\nimport { getAbsoluteUrl } from \"../../utils/url.ts\";\n```\n\n**Add webhook configuration variables after health data:**\n\n    // Check webhook configuration\n    const webhookSecret = Deno.env.get(\"NOTION_WEBHOOK_SECRET\");\n    const webhookConfigured = !!webhookSecret;\n    const webhookEndpoint = getAbsoluteUrl(c, '/tasks/notion-webhook');\n\n**Note:** We use `getAbsoluteUrl()` here because the webhook endpoint needs to be shared with external services (Notion), unlike internal links which should remain relative.\n\n**Add CSS styles for webhook section (after existing styles):**\n\n    .webhook-section {\n      background: #f8f9fa;\n      border: 1px solid #e9ecef;\n      border-radius: 6px;\n      padding: 20px;\n      margin-top: 20px;\n    }\n    .webhook-section h3 {\n      color: #333;\n      margin-top: 0;\n      margin-bottom: 15px;\n    }\n    .webhook-url {\n      margin-bottom: 20px;\n    }\n    .webhook-endpoint {\n      font-family: 'Monaco', 'Menlo', 'Ubuntu Mono', monospace;\n      font-size: 14px;\n      background: white;\n      padding: 12px;\n      border: 1px solid #dee2e6;\n      border-radius: 4px;\n      word-break: break-all;\n      user-select: all;\n    }\n    .test-section {\n      border-top: 1px solid #dee2e6;\n      padding-top: 20px;\n      margin-top: 20px;\n    }\n    .test-section p {\n      margin-bottom: 15px;\n      color: #495057;\n    }\n    .test-section code {\n      background: #f8f9fa;\n      padding: 2px 4px;\n      border-radius: 3px;\n      font-family: 'Monaco', 'Menlo', 'Ubuntu Mono', monospace;\n      font-size: 13px;\n    }\n    .test-controls {\n      display: flex;\n      gap: 10px;\n      align-items: center;\n      margin-bottom: 15px;\n    }\n    .test-btn {\n      background: #28a745;\n      color: white;\n      border: none;\n      padding: 8px 16px;\n      border-radius: 4px;\n      cursor: pointer;\n      font-size: 14px;\n      transition: background-color 0.2s;\n    }\n    .test-btn:hover {\n      background: #218838;\n    }\n    .test-btn:disabled {\n      background: #6c757d;\n      cursor: not-allowed;\n    }\n    .api-key-input {\n      padding: 6px 10px;\n      border: 1px solid #ced4da;\n      border-radius: 4px;\n      font-size: 14px;\n      width: 200px;\n    }\n    .test-result {\n      background: white;\n      border: 1px solid #dee2e6;\n      border-radius: 4px;\n      padding: 15px;\n      font-family: 'Monaco', 'Menlo', 'Ubuntu Mono', monospace;\n      font-size: 13px;\n      min-height: 60px;\n      white-space: pre-wrap;\n    }\n    .result-success {\n      border-color: #28a745;\n      background: #d4edda;\n    }\n    .result-error {\n      border-color: #dc3545;\n      background: #f8d7da;\n    }\n\n**Add simplified webhook section HTML (after health section):**\n\n    \n      Webhook Endpoint
\n\n      \n        ${webhookEndpoint}
\n      
\n\n      \n        To test: Get the NOTION_WEBHOOK_SECRET value from your vars.env file (provided by your administrator)
\n        \n          \n          \n        
\n        Ready to test webhook authentication...
\n      
\n    
\n\n**Add JavaScript for form submission (before closing body tag):**\n\n    \n\n## Testing Verification\n\n1. Test `/tasks/debug-webhook` (GET) - should show webhook configuration\n2. Test `/tasks/test` (POST) without header - should return 401\n3. Test `/tasks/test` (POST) with wrong key - should return 403\n4. Test `/tasks/test` (POST) with correct key - should return 200\n5. Verify user authentication still works for other routes\n6. Check dashboard shows simplified webhook testing interface\n7. Verify form submission works with both button click and Enter key\n\n## Expected Results\n\nMake sure these are true:\n\n- Webhook endpoints only accessible with correct X-API-KEY\n- User authentication preserved for non-webhook routes\n- Dashboard provides clean, simplified webhook testing interface\n- Form submission works via button click or Enter key\n- All authentication flows working independently\n- Webhook endpoint ready for Notion integration\n\n## Security Considerations\n\n- Webhook middleware should return early (success or failure) without calling\n  next()\n- Never pass webhook requests to the global auth middleware\n- Maintain complete separation between webhook auth and user auth mechanisms\n- Consider rate limiting on /tasks endpoints\n- Use generic error messages that don't reveal authentication mechanism details\n\n## Route Structure\n\n- /tasks/\\* → Webhook auth required (Notion POSTs)\n- /api/\\* → User auth required (if applicable)\n- /\\* → Public routes (static files, etc.)\n\n### 5\\. Update Documentation\n\nUpdate README.md to include webhook authentication section and environment\nvariables.\n\n## Final Deliverables\n\nMake sure these are true:\n\n- Webhook authentication middleware implemented\n- Middleware order correctly configured\n- Test endpoints available\n- Simplified dashboard with webhook testing form\n- Form supports both click and keyboard submission\n- Documentation updated\n- Ready for Notion webhook configuration\n\n---\n\n**Townie, stop here! Before proceeding to additional steps, confirm that this\nstep is working correctly.**\n\n**If everything is working as intended**: conclude this step with these\nmessages:\n\n1. Briefly summarize the work that was just completed and why it's important\n1. Provide the URL to the dashboard endpoint, from which the user can test the\n   new webhook endpoint with the webhook form\n1. Explain that the user will need the `NOTION_WEBHOOK_SECRET` from the\n   `vars.env` to test the webhook endpoint successfully\n1. Explain that we will implement webhooks from Notion in a step that comes\n   later\n",[],[154,199],{"_102":206,"_95":207,"_155":208,"_157":156,"_158":137,"_159":209,"_161":162,"_104":210,"_99":32,"_164":177,"_166":211,"_167":22,"_168":-5,"_169":212,"_98":213},"82c4257c-7a05-11f0-9236-0224a6c84d84","2-stage-questions-openai.md","82c4264e-7a05-11f0-9236-0224a6c84d84",8,["D",1755282453562],"# Instructions: Implementing Incremental Findings Analysis\n\n## Context\n\nYou have a cron job that processes research questions against multiple\ntranscripts using OpenAI, then updates a Notion report page. The original\nimplementation had two problems:\n\n1. Used non-existent OpenAI API endpoints\n2. Processed all questions in memory then updated Notion once at the end (bad\n   for 30+ questions)\n\nRough notes: does the openai service first ask a question of all transcripts,\ngenerate an answer per question, and then look for consensus across all 10\nanswers from all 10 transcripts? do not write code, just answer the question.\n\nYes, exactly! The OpenAI service uses a two-stage approach:\n\nStage 1 (Individual Analysis):\n\nTakes each transcript individually Asks the same research question to each\ntranscript separately Generates an individual analysis for each transcript that\nincludes: A stance (supports/mixed/contradicts/not enough evidence) Key findings\nSupporting quotes Rationale Confidence score Stage 2 (Consensus Synthesis):\n\nTakes all the individual analyses from Stage 1 Feeds them to a second OpenAI\ncall that acts as a \"senior research synthesizer\" This second call looks across\nall the individual answers to determine: Overall consensus level\n(high/medium/low) A direct answer that synthesizes all individual responses Key\npoints most transcripts agree on Notable disagreements or outliers Overall\nconfidence score So if you have 10 transcripts, it makes 10 individual OpenAI\ncalls in Stage 1, then 1 synthesis call in Stage 2 (total of 11 API calls per\nquestion). The final answer is based on finding patterns and consensus across\nall 10 individual transcript analyses.\n\n## Goal\n\nImplement a 2-stage OpenAI analysis with incremental Notion updates for\nreal-time progress visibility and fault tolerance.\n\n## Architecture & Separation of Concerns\n\n**This implementation follows Val Town's layered architecture:**\n\n- **Cron** (`findings-analysis.ts`): Simple trigger, calls controller\n- **Controller** (`findings.controller.ts`): Business logic orchestration,\n  environment variables, error handling\n- **Services** (`openai.service.ts`, `notion.service.ts`): Pure API calls with\n  consistent response format\n- **Types** (`findings.types.ts`): Shared interfaces\n\n**Key principle**: Services only make API calls and return\n`{ success, data?, error? }`. Controllers handle business logic and coordinate\nmultiple service calls.\n\n## Step 1: Rewrite OpenAI Service (`/backend/services/openai.service.ts`)\n\n**Replace entire file with 2-stage implementation:**\n\n### Core Types\n\n    export interface ConsensusAnalysisRequest {\n      question: string;\n      transcripts: string[];\n      apiKey: string;\n    }\n\n    export interface ConsensusAnswer {\n      question: string;\n      consensusLevel: \"high\" | \"medium\" | \"low\";\n      directAnswer: string;\n      keyPoints: string[];\n      disagreements: string[];\n      confidence: number; // 0..1\n      methodologyNote: string;\n      perTranscriptFindings: PerTranscriptFinding[];\n    }\n\n    export interface ConsensusAnalysisResponse {\n      success: boolean;\n      data?: ConsensusAnswer;\n      error?: string;\n      timestamp: string;\n    }\n\n### Stage 1: Individual Transcript Analysis\n\n**Function**:\n`analyzeIndividualTranscript(openai, model, question, transcript, transcriptKey, maxChars)`\n\n**Implementation details:**\n\n- Use `openai.chat.completions.create()` with actual OpenAI client\n- Model: `gpt-4o-mini` (cost efficient)\n- System prompt: \"You are a meticulous research analyst. Analyze the provided\n  interview transcript to answer a specific research question.\"\n- Response format: `{ type: \"json_object\" }`\n- Expected JSON structure:\n\n      {\n        \"stance\": \"supports|mixed|contradicts|not enough evidence\",\n        \"keyFindings\": [\"finding 1\", \"finding 2\"],\n        \"supportingQuotes\": [\"quote 1\", \"quote 2\"],\n        \"rationale\": \"Brief explanation\",\n        \"confidence\": 0.85\n      }\n\n- Error handling: Return fallback analysis with stance \"not enough evidence\" and\n  confidence 0.0\n- Truncate transcripts to maxChars (default 15,000) to avoid token limits\n\n### Stage 2: Consensus Synthesis\n\n**Function**: `synthesizeConsensus(openai, model, question, individualAnalyses)`\n\n**Implementation details:**\n\n- Use `openai.chat.completions.create()` with actual OpenAI client\n- Model: `gpt-4o` (high quality reasoning)\n- System prompt: \"You are a senior research synthesizer. You will receive\n  multiple individual analyses of different transcripts for the same research\n  question.\"\n- Input: Formatted string of all individual analyses\n- Response format: `{ type: \"json_object\" }`\n- Expected JSON structure:\n\n      {\n        \"consensusLevel\": \"high|medium|low\",\n        \"directAnswer\": \"Clear, direct answer\",\n        \"keyPoints\": [\"point 1\", \"point 2\"],\n        \"disagreements\": [\"disagreement 1\"],\n        \"confidence\": 0.85,\n        \"methodologyNote\": \"Brief explanation\"\n      }\n\n- Consensus level logic:\n  - HIGH: 80%+ of transcripts support same conclusion\n  - MEDIUM: 60-79% agreement\n  - LOW: <60% agreement or significant contradictions\n- Error handling: Create fallback consensus using stance distribution from Stage\n  1 results\n\n### Concurrency Control\n\n**Function**: `mapWithLimit(items, limit, fn)`\n\n- Process transcripts with configurable concurrency (default: 3)\n- Prevents API rate limiting and timeouts\n- Sequential processing within each transcript, parallel across transcripts\n\n### Main Export Function\n\n**Function**: `analyzeTranscriptsWithConsensus(request, options?)`\n\n**Flow:**\n\n1. Validate inputs (apiKey, question, transcripts)\n2. Create OpenAI client with timeout\n3. Stage 1: Process all transcripts in parallel (with limit) → individual\n   analyses\n4. Stage 2: Synthesize consensus from individual analyses → final answer\n5. Return `ConsensusAnalysisResponse` format\n\n**Error handling:**\n\n- Validation errors: Return immediately with descriptive error\n- Stage 1 errors: Continue with other transcripts, mark failed ones\n- Stage 2 errors: Use fallback consensus calculation\n- All errors include timestamp and preserve partial results\n\n## Step 2: Add New Notion Service Functions (`/backend/services/notion.service.ts`)\n\n**Add these functions following service pattern:**\n\n    export async function deletePageContent(pageId: string) {\n      try {\n        const existingBlocks = await notion.blocks.children.list({ block_id: pageId });\n        for (const block of existingBlocks.results) {\n          await notion.blocks.delete({ block_id: block.id });\n        }\n        return { success: true, data: { deletedBlocks: existingBlocks.results.length }, timestamp: new Date().toISOString() };\n      } catch (error) {\n        return { success: false, error: error.message, timestamp: new Date().toISOString() };\n      }\n    }\n\n    export async function appendPageContent(pageId: string, blocks: any[]) {\n      try {\n        const response = await notion.blocks.children.append({ block_id: pageId, children: blocks });\n        return { success: true, data: response, timestamp: new Date().toISOString() };\n      } catch (error) {\n        return { success: false, error: error.message, timestamp: new Date().toISOString() };\n      }\n    }\n\n**Service responsibilities:**\n\n- Pure API calls to Notion\n- Consistent `{ success, data?, error?, timestamp }` response format\n- No business logic or environment variable handling\n- Error handling only for API failures\n\n## Step 3: Restructure Controller (`/backend/controllers/findings.controller.ts`)\n\n**Controller responsibilities:**\n\n- Handle environment variables (`OPENAI_API_KEY`, database IDs)\n- Orchestrate multiple service calls\n- Implement business rules (incremental updates, error recovery)\n- Transform data between services\n- Comprehensive logging and error handling\n\n### Key Changes\n\n**1\\. Import new Notion functions:**\n\n    import { deletePageContent, appendPageContent } from \"../services/notion.service.ts\";\n\n**2\\. Early initialization (after finding report page):**\n\n    // Set status to \"Working...\" \n    const statusUpdateResult = await updatePageStatus(reportPageId, \"Working...\");\n\n    // Clear existing content (clean slate)\n    const clearContentResult = await deletePageContent(reportPageId);\n\n**3\\. Replace batch processing with incremental loop:**\n\n    // OLD: Collect all results, then update once\n    const analyses: EnhancedQuestionAnalysis[] = [];\n    for (const question of questions) {\n      // analyze and push to analyses array\n    }\n    const contentBlocks = formatEnhancedReportContent(analyses);\n    await updatePageContent(reportPageId, contentBlocks);\n\n    // NEW: Process and immediately append each question\n    for (const question of questions) {\n      const analysisResult = await analyzeTranscriptsWithConsensus({...});\n      if (analysisResult.success) {\n        const questionAnalysis = { question: question.name, consensusAnswer: analysisResult.data };\n        const contentBlocks = formatEnhancedReportContent([questionAnalysis]);\n        await appendPageContent(reportPageId, contentBlocks);\n      } else {\n        const errorBlocks = createErrorBlock(question.name, analysisResult.error);\n        await appendPageContent(reportPageId, errorBlocks);\n      }\n    }\n\n**4\\. Add error block creation:**\n\n    function createErrorBlock(questionName: string, errorMessage: string): any[] {\n      return [\n        { object: \"block\", type: \"heading_2\", heading_2: { rich_text: [{ type: \"text\", text: { content: questionName } }] } },\n        { object: \"block\", type: \"callout\", callout: { \n          icon: { emoji: \"❌\" }, \n          rich_text: [{ type: \"text\", text: { content: `Error: ${errorMessage}` }, annotations: { bold: true, color: \"red\" } }] \n        }},\n        { object: \"block\", type: \"divider\", divider: {} }\n      ];\n    }\n\n**5\\. Track success metrics:**\n\n    let questionsProcessed = 0;\n    let questionsSucceeded = 0;\n    // Increment in loop, return both in response\n\n**Business logic handled by controller:**\n\n- Environment variable validation\n- Status management (\"Not started\" → \"Working...\" → \"Done\")\n- Error recovery (continue processing other questions if one fails)\n- Progress tracking and logging\n- Data transformation between OpenAI and Notion formats\n\n## Step 4: Update Types (`/backend/types/findings.types.ts`)\n\n**Add to ProcessFindingsResponse:**\n\n    questionsSucceeded?: number;\n\n**Type responsibilities:**\n\n- Shared interfaces between layers\n- No business logic\n- Clear data contracts\n\n## Step 5: Update Cron Logging (`/backend/crons/findings-analysis.ts`)\n\n**Add logging for success metrics:**\n\n    if (result.success) {\n      console.log(`✅ Questions succeeded: ${result.questionsSucceeded}`);\n      if (result.questionsSucceeded !== result.questionsProcessed) {\n        console.log(`⚠️ Some questions failed: ${result.questionsProcessed - result.questionsSucceeded} errors`);\n      }\n    }\n\n**Cron responsibilities:**\n\n- Simple trigger mechanism\n- Basic logging of results\n- No business logic or error handling\n\n## Step 6: Test Implementation\n\n**Create temporary test endpoint:**\n\n    // /run-findings-cron.ts (HTTP)\n    import findingsAnalysisCron from \"./backend/crons/findings-analysis.ts\";\n    export default async function(req: Request) {\n      const result = await findingsAnalysisCron();\n      return new Response(JSON.stringify({ message: \"Cron execution completed\", result }, null, 2));\n    }\n\n**Test scenarios:**\n\n1. No syntax errors\n2. Handles \"No report pages found\" gracefully\n3. OpenAI service works with sample data\n4. Service response formats are consistent\n\n## Step 7: Clean Up\n\n**Remove test endpoint after verification**\n\n## Expected Outcome\n\n**Separation of concerns achieved:**\n\n- **Cron**: Minimal trigger\n- **Controller**: Business orchestration, environment handling, error recovery\n- **Services**: Pure API calls with consistent responses\n- **Types**: Clear data contracts\n\n**For 30 questions, users will see:**\n\n1. Report status changes to \"Working...\" immediately\n2. Questions appear in Notion one by one as they complete\n3. Failed questions show error blocks instead of breaking the entire process\n4. Final status shows \"Done\" with success metrics\n5. No timeouts or memory issues\n6. Fault tolerance - partial results preserved if process fails mid-way\n\n**Performance:**\n\n- ~70% cost reduction using gpt-4o-mini for Stage 1\n- Real-time progress visibility\n- Handles failures gracefully\n- Scales to 30+ questions without issues\n\n## Updated Instructions: Add Timing Capture\n\n### Step 3 Enhancement: Add Timing to Controller (`/backend/controllers/findings.controller.ts`)\n\n**Add timing capture at the start of processing:**\n\n    // After clearing page content, capture start time\n    const processingStartTime = new Date();\n    console.log(`⏱️ Starting question processing at ${processingStartTime.toISOString()}`);\n\n**Add timing capture after all questions are processed:**\n\n    // After the incremental processing loop, before final status update\n    const processingEndTime = new Date();\n    const totalTimeMs = processingEndTime.getTime() - processingStartTime.getTime();\n    const totalMinutes = Math.floor(totalTimeMs / 60000);\n    const totalSeconds = Math.floor((totalTimeMs % 60000) / 1000);\n\n    console.log(`⏱️ Completed question processing at ${processingEndTime.toISOString()}`);\n    console.log(`⏱️ Total processing time: ${totalMinutes}m ${totalSeconds}s`);\n\n    // Create timing summary block\n    const timingSummaryBlocks = createTimingSummaryBlock(totalMinutes, totalSeconds, questionsProcessed, questionsSucceeded);\n\n    // Append timing summary as final block\n    const timingAppendResult = await appendPageContent(reportPageId, timingSummaryBlocks);\n    if (!timingAppendResult.success) {\n      console.error(`⚠️ Failed to append timing summary: ${timingAppendResult.error}`);\n    }\n\n**Add timing summary block creation function:**\n\n    /**\n     * Create timing summary block for the end of the report\n     */\n    function createTimingSummaryBlock(minutes: number, seconds: number, totalQuestions: number, successfulQuestions: number): any[] {\n      const timeDisplay = minutes > 0 ? `${minutes}m ${seconds}s` : `${seconds}s`;\n      const successRate = Math.round((successfulQuestions / totalQuestions) * 100);\n      \n      return [\n        {\n          object: \"block\",\n          type: \"divider\",\n          divider: {}\n        },\n        {\n          object: \"block\",\n          type: \"heading_3\",\n          heading_3: {\n            rich_text: [\n              {\n                type: \"text\",\n                text: {\n                  content: \"📊 Report Generation Summary\"\n                }\n              }\n            ]\n          }\n        },\n        {\n          object: \"block\",\n          type: \"callout\",\n          callout: {\n            icon: { emoji: \"⏱️\" },\n            rich_text: [\n              {\n                type: \"text\",\n                text: {\n                  content: `Total time to generate report: ${timeDisplay}`\n                },\n                annotations: { bold: true }\n              }\n            ]\n          }\n        },\n        {\n          object: \"block\",\n          type: \"paragraph\",\n          paragraph: {\n            rich_text: [\n              {\n                type: \"text\",\n                text: {\n                  content: `Questions processed: ${successfulQuestions}/${totalQuestions} (${successRate}% success rate)`\n                }\n              }\n            ]\n          }\n        },\n        {\n          object: \"block\",\n          type: \"paragraph\",\n          paragraph: {\n            rich_text: [\n              {\n                type: \"text\",\n                text: {\n                  content: `Generated on: ${new Date().toLocaleString()}`\n                },\n                annotations: { italic: true, color: \"gray\" }\n              }\n            ]\n          }\n        }\n      ];\n    }\n\n### Updated Processing Flow\n\n**The complete timing flow:**\n\n1. Find report page → Set status to \"Working...\" → Clear content\n2. **Capture start time** ⏱️\n3. Process each question incrementally (append immediately)\n4. **Capture end time** ⏱️\n5. **Calculate and append timing summary** 📊\n6. Set status to \"Done\"\n\n### Expected Output\n\n**At the bottom of every report, users will see:**\n\n    📊 Report Generation Summary\n\n    ⏱️ Total time to generate report: 3m 45s\n\n    Questions processed: 28/30 (93% success rate)\n\n    Generated on: 8/15/2025, 3:45:23 PM\n\n**Benefits:**\n\n- ✅ Users can see how long the analysis took\n- ✅ Performance tracking for optimization\n- ✅ Success rate visibility\n- ✅ Clear completion timestamp\n- ✅ Professional report footer\n\n**Implementation notes:**\n\n- Time calculation excludes initial setup (finding pages, fetching data)\n- Focuses on actual question processing time\n- Handles both successful and failed questions in timing\n- Timing block is always appended, even if some questions fail",[],[154,207],{"_102":215,"_95":216,"_155":217,"_157":156,"_158":137,"_159":218,"_161":162,"_104":219,"_99":32,"_164":177,"_166":220,"_167":22,"_168":-5,"_169":221,"_98":222},"fdca3d90-841b-11f0-bd95-0224a6c84d84","email-PDF-attachments.md","fdca3e3a-841b-11f0-bd95-0224a6c84d84",38,["D",1756391620516],"Complete Instructions: Email Handler for PDF Attachments to Notion\n==================================================================\n\nObjective\n---------\n\nCreate an email trigger that processes PDF attachments from incoming emails and uploads them to a Notion database, with each PDF getting its own page. Use Notion's official 3-step file upload API and maintain proper separation of concerns throughout the Val architecture.\n\nPrerequisites\n-------------\n\n*   `FINDINGS_TRANSCRIPTS_DB_ID` environment variable set to target Notion database ID\n*   `NOTION_API_KEY` environment variable set for Notion API authentication\n*   Notion database with properties: Name (title), Filename (rich text), PDF (files), Status (select)\n\nImplementation Steps\n--------------------\n\n### 1\\. Create Email Types (`/backend/types/email.types.ts`)\n\n    /**\n     * Email Processing Type Definitions\n     * TypeScript interfaces for email attachment processing\n     */\n    \n    // Val Town's actual Email interface\n    export interface Email {\n      from: string;\n      to: string[];\n      cc?: string | string[];\n      bcc?: string | string[];\n      subject?: string;\n      text?: string;\n      html?: string;\n      attachments?: File[];  // File[] array, NOT custom objects\n    }\n    \n    export interface ProcessEmailAttachmentsRequest {\n      attachments: File[];  // Use actual File objects\n      databaseId: string;\n      pdfPropertyName?: string;\n    }\n    \n    export interface AttachmentProcessingResult {\n      filename: string;\n      pageCreated: boolean;\n      pageId?: string;\n      fileUploaded: boolean;\n      uploadId?: string;\n      error?: string;\n    }\n    \n    export interface ProcessEmailAttachmentsResponse {\n      success: boolean;\n      totalAttachments: number;\n      processedAttachments: number;\n      successfulUploads: number;\n      results: AttachmentProcessingResult[];\n      errors: string[];\n      timestamp: string;\n    }\n    \n    export interface NotionFileUploadObject {\n      id: string;\n      upload_url: string;\n      expiry_time: string;\n    }\n    \n\n### 2\\. Add Notion File Upload Services (`/backend/services/notion.service.ts`)\n\nAdd these three functions to the existing service file:\n\n    /**\n     * Step 1: Create File Upload Object\n     * Creates an upload object in Notion and returns id and upload_url\n     */\n    export async function createNotionFileUpload() {\n      try {\n        const apiKey = Deno.env.get(\"NOTION_API_KEY\");\n        \n        const createUploadResponse = await fetch('https://api.notion.com/v1/file_uploads', {\n          method: 'POST',\n          headers: {\n            'Authorization': `Bearer ${apiKey}`,\n            'Content-Type': 'application/json',\n            'Notion-Version': '2022-06-28'\n          },\n          body: JSON.stringify({})\n        });\n    \n        if (!createUploadResponse.ok) {\n          const errorText = await createUploadResponse.text();\n          throw new Error(`HTTP ${createUploadResponse.status}: ${errorText}`);\n        }\n    \n        const uploadObject = await createUploadResponse.json();\n        \n        return {\n          success: true,\n          data: uploadObject,\n          timestamp: new Date().toISOString(),\n        };\n      } catch (error) {\n        return {\n          success: false,\n          error: error.message,\n          timestamp: new Date().toISOString(),\n        };\n      }\n    }\n    \n    /**\n     * Step 2: Upload File Contents\n     * Uploads the actual file content to Notion's upload URL using FormData\n     */\n    export async function uploadFileToNotionUrl(uploadUrl: string, pdfFile: File) {\n      try {\n        const apiKey = Deno.env.get(\"NOTION_API_KEY\");\n        \n        const formData = new FormData();\n        formData.append('file', pdfFile);\n        \n        const uploadResponse = await fetch(uploadUrl, {\n          method: 'POST',\n          headers: {\n            'Authorization': `Bearer ${apiKey}`,\n            'Notion-Version': '2022-06-28'\n            // DO NOT set Content-Type - let FormData handle it\n          },\n          body: formData\n        });\n    \n        if (!uploadResponse.ok) {\n          const errorText = await uploadResponse.text();\n          throw new Error(`HTTP ${uploadResponse.status}: ${errorText}`);\n        }\n    \n        const uploadResult = await uploadResponse.json();\n        \n        return {\n          success: true,\n          data: uploadResult,\n          timestamp: new Date().toISOString(),\n        };\n      } catch (error) {\n        return {\n          success: false,\n          error: error.message,\n          timestamp: new Date().toISOString(),\n        };\n      }\n    }\n    \n    /**\n     * Step 3: Attach File to Page\n     * Attaches the uploaded file to a Notion page property using file_upload type\n     */\n    export async function attachFileUploadToPage(pageId: string, propertyName: string, uploadObjectId: string, fileName: string) {\n      try {\n        const apiKey = Deno.env.get(\"NOTION_API_KEY\");\n        \n        const attachResponse = await fetch(`https://api.notion.com/v1/pages/${pageId}`, {\n          method: 'PATCH',\n          headers: {\n            'Authorization': `Bearer ${apiKey}`,\n            'Notion-Version': '2022-06-28',\n            'Content-Type': 'application/json'\n          },\n          body: JSON.stringify({\n            properties: {\n              [propertyName]: {\n                files: [{\n                  name: fileName,\n                  type: \"file_upload\",  // NOT \"file\" or \"external\"\n                  file_upload: {\n                    id: uploadObjectId  // ID from Step 1\n                  }\n                }]\n              }\n            }\n          })\n        });\n    \n        if (!attachResponse.ok) {\n          const errorText = await attachResponse.text();\n          throw new Error(`HTTP ${attachResponse.status}: ${errorText}`);\n        }\n    \n        const attachResult = await attachResponse.json();\n        \n        return {\n          success: true,\n          data: attachResult,\n          timestamp: new Date().toISOString(),\n        };\n      } catch (error) {\n        return {\n          success: false,\n          error: error.message,\n          timestamp: new Date().toISOString(),\n        };\n      }\n    }\n    \n\n### 3\\. Create Email Utilities (`/backend/utils/email.utils.ts`)\n\n    /**\n     * Email Utility Functions\n     * Utility functions for processing email data and attachments\n     */\n    \n    /**\n     * Extract interviewee name from filename\n     * Looks for content in parentheses, falls back to filename without extension\n     * \n     * @param filename - The original filename from email attachment\n     * @returns Extracted interviewee name or filename without extension\n     * \n     * @example\n     * extractIntervieweeName(\"Dealfront discussion (Anna Moch) - 2025_08_04.pdf\")\n     * // Returns: \"Anna Moch\"\n     * \n     * extractIntervieweeName(\"meeting-notes.pdf\")\n     * // Returns: \"meeting-notes\"\n     */\n    export function extractIntervieweeName(filename: string): string {\n      // Look for content in parentheses\n      const match = filename.match(/\\(([^)]+)\\)/);\n      if (match) {\n        return match[1].trim(); // Return content inside first parentheses\n      }\n      \n      // Fallback: use filename without extension\n      return filename.replace(/\\.[^/.]+$/, \"\").trim();\n    }\n    \n\n### 4\\. Create Email Controller (`/backend/controllers/emails.controller.ts`)\n\n    /**\n     * Emails Controller\n     * Business logic for processing email PDF attachments to Notion\n     */\n    \n    import {\n      createPageInDatabase,\n      createNotionFileUpload,\n      uploadFileToNotionUrl,\n      attachFileUploadToPage,\n      updatePageProperties,\n    } from \"../services/notion.service.ts\";\n    import { extractIntervieweeName } from \"../utils/email.utils.ts\";\n    import type {\n      ProcessEmailAttachmentsRequest,\n      ProcessEmailAttachmentsResponse,\n      AttachmentProcessingResult,\n      NotionFileUploadObject,\n    } from \"../types/email.types.ts\";\n    \n    /**\n     * Main function to process email PDF attachments\n     * Creates child pages in Notion database and uploads PDFs using 3-step process\n     */\n    export async function processEmailPdfAttachments(\n      request: ProcessEmailAttachmentsRequest\n    ): Promise {\n      const timestamp = new Date().toISOString();\n      const results: AttachmentProcessingResult[] = [];\n      const errors: string[] = [];\n      const pdfPropertyName = request.pdfPropertyName || \"PDF\";\n    \n      console.log(`📧 Processing ${request.attachments.length} email attachments`);\n    \n      // Filter for PDF attachments only using File object properties\n      const pdfAttachments = request.attachments.filter(\n        file => file.name.toLowerCase().endsWith('.pdf') || \n                file.type.includes('pdf')\n      );\n    \n      console.log(`📄 Found ${pdfAttachments.length} PDF attachments`);\n    \n      if (pdfAttachments.length === 0) {\n        return {\n          success: true,\n          totalAttachments: request.attachments.length,\n          processedAttachments: 0,\n          successfulUploads: 0,\n          results: [],\n          errors: [\"No PDF attachments found\"],\n          timestamp,\n        };\n      }\n    \n      // Process each PDF attachment - each gets its own page\n      for (const file of pdfAttachments) {\n        const result = await processSinglePdfFile(\n          file,\n          request.databaseId,\n          pdfPropertyName\n        );\n        \n        results.push(result);\n        \n        if (result.error) {\n          errors.push(`${file.name}: ${result.error}`);\n        }\n    \n        // Small delay between attachments to avoid rate limiting\n        await new Promise(resolve => setTimeout(resolve, 500));\n      }\n    \n      const successfulUploads = results.filter(r => r.fileUploaded).length;\n      const processedAttachments = results.filter(r => r.pageCreated).length;\n    \n      console.log(`📊 Processing complete: ${processedAttachments}/${pdfAttachments.length} pages created, ${successfulUploads}/${pdfAttachments.length} files uploaded`);\n    \n      return {\n        success: errors.length === 0,\n        totalAttachments: request.attachments.length,\n        processedAttachments,\n        successfulUploads,\n        results,\n        errors,\n        timestamp,\n      };\n    }\n    \n    /**\n     * Process a single PDF File using two-phase approach\n     * Phase 1: Create page in database\n     * Phase 2: Upload file using 3-step Notion API\n     */\n    async function processSinglePdfFile(\n      file: File,\n      databaseId: string,\n      pdfPropertyName: string\n    ): Promise {\n      const result: AttachmentProcessingResult = {\n        filename: file.name,\n        pageCreated: false,\n        fileUploaded: false,\n      };\n    \n      try {\n        console.log(`📄 Processing PDF: ${file.name}`);\n        console.log(`   - File type: ${file.type}`);\n        console.log(`   - File size: ${file.size} bytes`);\n    \n        // Extract interviewee name from filename\n        const intervieweeName = extractIntervieweeName(file.name);\n        console.log(`   - Extracted interviewee name: \"${intervieweeName}\"`);\n    \n        // Phase 1: Create page in database\n        const pageProperties = {\n          \"Name\": {\n            title: [\n              {\n                text: {\n                  content: intervieweeName\n                }\n              }\n            ]\n          },\n          \"Filename\": {\n            rich_text: [\n              {\n                text: {\n                  content: file.name\n                }\n              }\n            ]\n          },\n          \"Status\": {\n            select: {\n              name: \"Uploading...\"\n            }\n          }\n        };\n    \n        const pageResult = await createPageInDatabase(databaseId, pageProperties);\n        \n        if (!pageResult.success) {\n          result.error = `Failed to create page: ${pageResult.error}`;\n          return result;\n        }\n    \n        result.pageCreated = true;\n        result.pageId = pageResult.data.id;\n        console.log(`✅ Page created: ${result.pageId}`);\n    \n        // Phase 2: Upload file using 3-step process\n        const uploadResult = await uploadPdfToNotionPage(\n          result.pageId,\n          file,\n          pdfPropertyName\n        );\n    \n        if (uploadResult.success) {\n          result.fileUploaded = true;\n          result.uploadId = uploadResult.uploadId;\n          console.log(`✅ File uploaded successfully: ${file.name}`);\n          \n          // Update status to \"Uploaded\"\n          const statusUpdateResult = await updatePageProperties(result.pageId, {\n            \"Status\": {\n              select: { name: \"Uploaded\" }\n            }\n          });\n          \n          if (statusUpdateResult.success) {\n            console.log(`✅ Status updated to \"Uploaded\" for ${file.name}`);\n          } else {\n            console.log(`⚠️ Failed to update status to \"Uploaded\": ${statusUpdateResult.error}`);\n          }\n          \n        } else {\n          result.error = `File upload failed: ${uploadResult.error}`;\n          console.log(`⚠️ File upload failed for ${file.name}, updating status to \"Upload error\"`);\n          \n          // Update status to \"Upload error\"\n          const statusUpdateResult = await updatePageProperties(result.pageId, {\n            \"Status\": {\n              select: { name: \"Upload error\" }\n            }\n          });\n          \n          if (statusUpdateResult.success) {\n            console.log(`✅ Status updated to \"Upload error\" for ${file.name}`);\n          } else {\n            console.log(`⚠️ Failed to update status to \"Upload error\": ${statusUpdateResult.error}`);\n          }\n        }\n    \n      } catch (error) {\n        result.error = `Unexpected error: ${error.message}`;\n        console.error(`❌ Error processing ${file.name}:`, error);\n      }\n    \n      return result;\n    }\n    \n    /**\n     * Upload PDF to Notion page using exact 3-step file upload process\n     */\n    async function uploadPdfToNotionPage(\n      pageId: string,\n      pdfFile: File,\n      propertyName: string\n    ): Promise<{ success: boolean; uploadId?: string; error?: string }> {\n      try {\n        // Validate file before upload\n        console.log(`📋 Validating file: ${pdfFile.name}`);\n        console.log(`   - Content type: ${pdfFile.type}`);\n        console.log(`   - File size: ${pdfFile.size} bytes`);\n        \n        if (pdfFile.size === 0) {\n          return { success: false, error: 'File is empty' };\n        }\n        \n        if (pdfFile.size > 50 * 1024 * 1024) { // 50MB limit\n          return { success: false, error: 'File too large (>50MB)' };\n        }\n    \n        // Step 1: Create File Upload Object\n        console.log(`📤 Step 1: Creating upload object for ${pdfFile.name}`);\n        const createUploadResult = await createNotionFileUpload();\n        \n        if (!createUploadResult.success) {\n          return { success: false, error: `Step 1 failed: ${createUploadResult.error}` };\n        }\n    \n        const uploadObject = createUploadResult.data;\n        console.log(`✅ Step 1 complete: Upload object created with ID ${uploadObject.id}`);\n    \n        // Step 2: Upload File Contents\n        console.log(`📤 Step 2: Uploading file contents for ${pdfFile.name}`);\n        \n        const uploadResult = await uploadFileToNotionUrl(uploadObject.upload_url, pdfFile);\n        \n        if (!uploadResult.success) {\n          return { success: false, error: `Step 2 failed: ${uploadResult.error}` };\n        }\n    \n        console.log(`✅ Step 2 complete: File contents uploaded`);\n        console.log(`   - Upload status: ${uploadResult.data.status}`);\n    \n        // Step 3: Attach File to Page\n        console.log(`📤 Step 3: Attaching file to page ${pageId}`);\n        const attachResult = await attachFileUploadToPage(\n          pageId,\n          propertyName,\n          uploadObject.id,  // Use the upload object ID from Step 1\n          pdfFile.name\n        );\n    \n        if (!attachResult.success) {\n          return { success: false, error: `Step 3 failed: ${attachResult.error}` };\n        }\n    \n        console.log(`✅ Step 3 complete: File attached to page with type \"file_upload\"`);\n        console.log(`   - Property updated: ${propertyName}`);\n    \n        return { success: true, uploadId: uploadObject.id };\n    \n      } catch (error) {\n        console.error(`❌ Upload process error:`, error);\n        return { success: false, error: `Upload process failed: ${error.message}` };\n      }\n    }\n    \n\n### 5\\. Create Email Handler (`/backend/emails/pdf-attachments.ts`)\n\n    /**\n     * Email Handler: PDF Attachments to Notion\n     * Processes incoming emails with PDF attachments and saves them to Notion database\n     * Each PDF attachment gets its own page in the database\n     */\n    \n    import { processEmailPdfAttachments } from \"../controllers/emails.controller.ts\";\n    import type { Email } from \"../types/email.types.ts\";\n    \n    export default async function(email: Email) {\n      try {\n        console.log(`📧 Received email: ${email.subject || 'No Subject'}`);\n        console.log(`📧 From: ${email.from}`);\n        console.log(`📧 Attachments: ${email.attachments?.length || 0}`);\n    \n        // Get database ID from environment\n        const databaseId = Deno.env.get('FINDINGS_TRANSCRIPTS_DB_ID');\n        if (!databaseId) {\n          console.error('❌ FINDINGS_TRANSCRIPTS_DB_ID environment variable not set');\n          return {\n            success: false,\n            error: 'Database ID not configured',\n            timestamp: new Date().toISOString(),\n          };\n        }\n    \n        // Check if email has attachments\n        if (!email.attachments || email.attachments.length === 0) {\n          console.log('📧 No attachments found in email');\n          return {\n            success: true,\n            message: 'No attachments to process',\n            timestamp: new Date().toISOString(),\n          };\n        }\n    \n        // Log attachment details for debugging\n        console.log(`📎 Processing ${email.attachments.length} attachments:`);\n        email.attachments.forEach((file, index) => {\n          console.log(`   ${index + 1}. ${file.name} (${file.type}, ${file.size} bytes)`);\n        });\n    \n        // Filter for PDF attachments\n        const pdfAttachments = email.attachments.filter(\n          file => file.name.toLowerCase().endsWith('.pdf') || \n                  file.type.includes('pdf')\n        );\n    \n        console.log(`📄 Found ${pdfAttachments.length} PDF attachments`);\n    \n        if (pdfAttachments.length === 0) {\n          console.log('📧 No PDF attachments found in email');\n          return {\n            success: true,\n            message: 'No PDF attachments to process',\n            timestamp: new Date().toISOString(),\n          };\n        }\n    \n        // Process attachments through controller\n        const result = await processEmailPdfAttachments({\n          attachments: email.attachments, // Pass all attachments, controller will filter PDFs\n          databaseId,\n          pdfPropertyName: 'PDF',\n        });\n    \n        // Log results\n        console.log(`📊 Email processing complete:`);\n        console.log(`   - Total attachments: ${result.totalAttachments}`);\n        console.log(`   - PDF attachments: ${pdfAttachments.length}`);\n        console.log(`   - Pages created: ${result.processedAttachments}`);\n        console.log(`   - Files uploaded: ${result.successfulUploads}`);\n        \n        if (result.errors.length > 0) {\n          console.log(`   - Errors: ${result.errors.length}`);\n          result.errors.forEach(error => console.log(`     • ${error}`));\n        }\n    \n        return {\n          success: result.success,\n          message: `Processed ${result.processedAttachments} PDF attachments, ${result.successfulUploads} uploaded successfully`,\n          data: {\n            totalAttachments: result.totalAttachments,\n            pdfAttachments: pdfAttachments.length,\n            processedAttachments: result.processedAttachments,\n            successfulUploads: result.successfulUploads,\n            results: result.results,\n          },\n          errors: result.errors,\n          timestamp: result.timestamp,\n        };\n    \n      } catch (error) {\n        console.error('❌ Email handler error:', error);\n        return {\n          success: false,\n          error: error.message,\n          timestamp: new Date().toISOString(),\n        };\n      }\n    }\n    \n\n### 6\\. Set Email Trigger\n\nUse `change_val_type` tool to configure `/backend/emails/pdf-attachments.ts` as email trigger.\n\n### 7\\. Create Directory READMEs\n\nCreate `/backend/emails/README.md`:\n\n    # Emails\n    \n    Email trigger handlers for processing incoming emails.\n    \n    ## Files\n    \n    - `pdf-attachments.ts` - Processes emails with PDF attachments and uploads them to Notion database pages\n    \n    ## Architecture\n    \n    Email handlers follow the same separation of concerns as other parts of the application:\n    \n    - **Email Handlers**: Extract data from email objects and call controllers\n    - **Controllers**: Business logic for processing email data\n    - **Services**: Pure API calls to external systems (Notion, etc.)\n    \n    ## PDF Attachments Handler\n    \n    The `pdf-attachments.ts` handler:\n    \n    1. Receives emails with PDF attachments\n    2. Creates child pages in the Notion database specified by `FINDINGS_TRANSCRIPTS_DB_ID`\n    3. Uploads each PDF to its own page using Notion's 3-step file upload API\n    4. Continues processing even if individual uploads fail\n    5. Returns detailed results for monitoring\n    \n    ### Page Properties\n    \n    Each PDF attachment creates a page with:\n    \n    - **Name**: Extracted interviewee name from filename (content in parentheses)\n    - **Filename**: Complete original filename from email attachment\n    - **PDF**: Uploaded file using Notion's file upload API\n    - **Status**: Upload progress tracking\n    \n    ### Status Progression\n    \n    - **\"Uploading...\"**: Initial status when page is created\n    - **\"Uploaded\"**: File successfully uploaded and attached to page\n    - **\"Upload error\"**: File upload failed (page still exists for retry)\n    \n    ### Filename Processing\n    \n    The handler extracts interviewee names using pattern matching:\n    \n    - **Pattern**: `(Name)` - Content inside first parentheses becomes the page title\n    - **Example**: `\"Interview (John Doe) - 2025_01_01.pdf\"` → Name: `\"John Doe\"`\n    - **Fallback**: If no parentheses found, uses filename without extension\n    \n    ### Environment Variables\n    \n    - `FINDINGS_TRANSCRIPTS_DB_ID` - Notion database ID where pages will be created\n    - `NOTION_API_KEY` - Notion API key for authentication\n    \n    ### Process Flow\n    \n    1. **Email Reception**: Handler receives email with attachments\n    2. **PDF Filtering**: Only processes PDF attachments (by extension or MIME type)\n    3. **Page Creation**: Creates a new page in the target database for each PDF\n    4. **File Upload**: Uses Notion's 3-step upload process:\n       - Create upload object\n       - Upload file contents\n       - Attach file to page\n    5. **Status Updates**: Updates page status based on upload success/failure\n    6. **Error Handling**: Continues processing even if individual steps fail\n    7. **Response**: Returns detailed processing results\n    \n\n### 8\\. Update Existing READMEs\n\nUpdate `/backend/README.md` to include `/emails` directory:\n\n    - `/emails` - Email trigger handlers for processing incoming emails\n    \n\nUpdate `/backend/controllers/README.md` to include emails controller:\n\n    - `emails.controller.ts` - Handles email PDF attachment processing workflow\n      - **PDF Processing**: Processes each PDF attachment as separate database page\n      - **Filename Parsing**: Extracts interviewee names from filenames using parentheses pattern\n      - **Status Tracking**: Manages upload status progression (\"Uploading...\" → \"Uploaded\" → \"Upload error\")\n      - **3-Step Upload**: Orchestrates Notion's file upload process (create → upload → attach)\n    \n\nUpdate `/backend/services/README.md` to include file upload functions:\n\n    - **File Upload API**: 3-step file upload process implementation\n      - `createNotionFileUpload()` - Step 1: Create upload object\n      - `uploadFileToNotionUrl()` - Step 2: Upload file contents via FormData\n      - `attachFileUploadToPage()` - Step 3: Attach file to page with \"file_upload\" type\n    \n\nUpdate `/backend/types/README.md` to include email types:\n\n    - `email.types.ts` - Email processing and attachment handling types\n      - **Email Interface**: Val Town's standard Email interface with File[] attachments\n      - **Processing Types**: Request/response interfaces for email attachment workflows\n      - **Result Types**: Detailed processing results and error tracking\n    \n\nUpdate `/backend/utils/README.md` to include email utils:\n\n    - `email.utils.ts` - Email processing utilities (filename parsing, interviewee name extraction)\n    \n\n### 9\\. Update Notion Types\n\nAdd file property type to `/backend/types/notion.types.ts`:\n\n    export interface NotionFileProperty {\n      files: Array<{\n        name: string;\n        type: \"file_upload\";\n        file_upload: {\n          id: string;\n        };\n      }>;\n    }\n    \n\nAnd update the main interface:\n\n    export interface NotionProperties {\n      [key: string]: NotionDateProperty | NotionSelectProperty | NotionRichTextProperty | NotionNumberProperty | NotionFileProperty;\n    }\n    \n\nCritical Implementation Details\n-------------------------------\n\n### Val Town Email Interface\n\n*   **MUST USE** `File[]` for attachments, NOT custom objects\n*   Use `file.name`, `file.type`, `file.size` properties\n*   **DO NOT** use `filename`, `contentType`, `content` properties\n\n### Notion 3-Step Upload Process\n\n*   **Step 1**: Create upload object with empty JSON body\n*   **Step 2**: Use FormData with File object, DO NOT set Content-Type header\n*   **Step 3**: Use `type: \"file_upload\"` with upload object ID\n\n### Separation of Concerns\n\n*   **Services**: Pure API calls only, return consistent format\n*   **Controllers**: Business logic orchestration, coordinate services\n*   **Utils**: Pure utility functions, domain-organized\n*   **Handlers**: Request/response handling, call controllers only\n\n### Status Management\n\n*   **\"Uploading...\"**: Set on page creation\n*   **\"Uploaded\"**: Set after successful 3-step upload\n*   **\"Upload error\"**: Set if upload fails (page still exists)\n\n### Error Handling\n\n*   Continue processing even if individual uploads fail\n*   Log detailed information for debugging\n*   Update status appropriately for failed uploads\n*   Return comprehensive results for monitoring\n\nExpected Behavior\n-----------------\n\nWhen email received with PDF attachments:\n\n1.  Creates separate Notion page for each PDF\n2.  Extracts interviewee name from filename for page title\n3.  Stores complete filename in Filename property\n4.  Uploads PDF using official Notion API\n5.  Updates status based on upload success/failure\n6.  Returns detailed processing results\n7.  Continues processing even if individual files fail\n\nThis implementation maintains proper separation of concerns, uses Val Town's actual email interface, implements Notion's official file upload API correctly, and provides comprehensive error handling and status tracking.",[],[154,216],{"_102":224,"_95":225,"_155":226,"_157":156,"_158":137,"_159":227,"_161":162,"_104":228,"_99":32,"_164":177,"_166":229,"_167":22,"_168":-5,"_169":230,"_98":231},"c8fa0d2a-79c6-11f0-95e0-0224a6c84d84","pdf-to-blob.md","c8fa0dd4-79c6-11f0-95e0-0224a6c84d84",6,["D",1755255513063],"# Complete Instructions for Creating the `/tasks/blob` Endpoint\n\n## Overview\n\nCreate a webhook endpoint that processes PDF text extraction and blob storage\nfor Notion pages, following the established separation of concerns architecture.\n\n## Prerequisites\n\nBefore starting, verify these files exist and understand their current state:\n\n- `/backend/routes/tasks/_tasks.routes.ts` (route definitions)\n- `/backend/controllers/tasks.controller.ts` (business logic)\n- `/backend/services/notion.service.ts` (Notion API calls)\n- `/backend/utils/pdfExtractor.ts` (PDF processing utilities)\n- `/backend/types/` (type definitions directory)\n\n## Step 1: Create Type Definitions\n\nCreate or update `/backend/types/tasks.types.ts` with the following content:\n\n    /**\n     * Task processing type definitions\n     */\n\n    export interface ProcessBlobRequest {\n      pageId: string;\n    }\n\n    export interface ProcessBlobResponse {\n      success: boolean;\n      pageId: string;\n      blobKey?: string;\n      textLength?: number;\n      statusUpdated?: boolean;\n      timestamp: string;\n      error?: string;\n      details?: string;\n    }\n\n    export interface WebhookPayload {\n      data: {\n        id: string;\n        [key: string]: any;\n      };\n      [key: string]: any;\n    }\n\n    export interface NotionPageProperties {\n      id: string;\n      properties: {\n        PDF?: {\n          type: 'files';\n          files: Array<{\n            type: 'external' | 'file';\n            external?: { url: string };\n            file?: { url: string };\n          }>;\n        };\n        [key: string]: any;\n      };\n    }\n\n## Step 2: Add Required Imports to Route File\n\nIn `/backend/routes/tasks/_tasks.routes.ts`, ensure these imports are present at\nthe top:\n\n    import { Hono } from \"npm:hono@3.12.12\";\n    import { processBlobExtraction, extractPageIdFromWebhook } from \"../../controllers/tasks.controller.ts\";\n\n## Step 3: Add Route Handler\n\nIn `/backend/routes/tasks/_tasks.routes.ts`, add this endpoint **before** the\ncatch-all handler (the `app.post(\"*\", ...)` route):\n\n    // PDF text extraction and blob storage endpoint\n    app.post(\"/blob\", async (c) => {\n      try {\n        // At this point, webhook auth has already passed\n        const body = await c.req.json();\n        console.log(\"📥 Blob processing webhook received:\", body);\n\n        // Extract page ID from webhook payload\n        const pageId = extractPageIdFromWebhook(body);\n        if (!pageId) {\n          return c.json({ error: \"Page ID is required\" }, 400);\n        }\n\n        // Process the blob extraction using controller\n        const result = await processBlobExtraction({ pageId });\n\n        // Return appropriate HTTP status based on result\n        if (result.success) {\n          return c.json(result);\n        } else {\n          const statusCode = result.error?.includes(\"No PDF\") || result.error?.includes(\"No text\") ? 400 : 500;\n          return c.json(result, statusCode);\n        }\n\n      } catch (error) {\n        console.error(\"❌ Error in blob endpoint:\", error);\n        return c.json({ \n          error: \"Internal server error\", \n          details: error.message,\n          timestamp: new Date().toISOString()\n        }, 500);\n      }\n    });\n\n## Step 4: Update Available Endpoints List\n\nIn the same file, find the catch-all handler and update the `availableEndpoints`\narray to include the new endpoint:\n\n    app.post(\"*\", (c) => {\n      const path = c.req.path;\n      const method = c.req.method;\n      const returnObj = {\n        error: \"Endpoint not found\",\n        path: path,\n        method: method,\n        availableEndpoints: [\n          \"GET /tasks/debug-webhook\",\n          \"POST /tasks/test\",\n          \"POST /tasks/notion-webhook\",\n          \"POST /tasks/blob\",  // <- Add this line\n        ],\n      };\n      console.log(returnObj);\n\n      return c.json(returnObj, 404);\n    });\n\n## Step 5: Add Required Imports to Controller\n\nIn `/backend/controllers/tasks.controller.ts`, ensure these imports are present\nat the top:\n\n    import { blob } from \"https://esm.town/v/std/blob\";\n    import { getPageProperties, updatePageStatus } from \"../services/notion.service.ts\";\n    import { extractTextFromPDFUrl } from \"../utils/pdfExtractor.ts\";\n    import type { \n      ProcessBlobRequest, \n      ProcessBlobResponse, \n      WebhookPayload, \n      NotionPageProperties \n    } from \"../types/tasks.types.ts\";\n\n## Step 6: Add Main Controller Function\n\nIn `/backend/controllers/tasks.controller.ts`, add this function:\n\n    /**\n     * Process PDF extraction and blob storage for a Notion page\n     */\n    export async function processBlobExtraction(request: ProcessBlobRequest): Promise {\n      const { pageId } = request;\n      const timestamp = new Date().toISOString();\n\n      try {\n        console.log(`🔍 Processing page: ${pageId}`);\n\n        // Get page properties from Notion\n        const pageResult = await getPageProperties(pageId);\n        if (!pageResult.success) {\n          console.error(`❌ Failed to get page properties: ${pageResult.error}`);\n          return {\n            success: false,\n            pageId,\n            timestamp,\n            error: `Failed to get page properties: ${pageResult.error}`\n          };\n        }\n\n        const page = pageResult.data;\n        console.log(`📄 Retrieved page properties for: ${page.id}`);\n\n        // Extract PDF URL from page properties\n        const pdfUrl = extractPdfUrl(page);\n        if (!pdfUrl) {\n          console.error(\"❌ No valid PDF found on page\");\n          return {\n            success: false,\n            pageId,\n            timestamp,\n            error: \"No PDF files found in PDF property\"\n          };\n        }\n\n        console.log(`📎 Found PDF URL: ${pdfUrl}`);\n\n        // Extract text from PDF\n        console.log(\"🔄 Starting PDF text extraction...\");\n        const extractedText = await extractTextFromPDFUrl(pdfUrl);\n        \n        if (!extractedText || extractedText.trim().length === 0) {\n          console.error(\"❌ No text extracted from PDF\");\n          return {\n            success: false,\n            pageId,\n            timestamp,\n            error: \"No text could be extracted from PDF\"\n          };\n        }\n\n        // Save extracted text to blob storage\n        const blobKey = `findings--transcripts--${pageId}`;\n        console.log(`💾 Saving extracted text to blob with key: ${blobKey}`);\n        \n        await blob.setJSON(blobKey, {\n          pageId: pageId,\n          extractedText: extractedText,\n          extractedAt: timestamp,\n          textLength: extractedText.length,\n          pdfUrl: pdfUrl\n        });\n\n        console.log(`✅ Text saved to blob storage successfully`);\n\n        // Update page status to \"Done\" and save blob key\n        console.log(\"🔄 Updating page status to 'Done' and saving blob key...\");\n        const statusResult = await updatePageStatus(pageId, \"Done\", blobKey);\n        if (!statusResult.success) {\n          console.error(`⚠️ Failed to update page status: ${statusResult.error}`);\n          // Don't fail the entire operation if status update fails\n        } else {\n          console.log(\"✅ Page status updated to 'Done' and blob key saved\");\n        }\n\n        return {\n          success: true,\n          pageId,\n          blobKey,\n          textLength: extractedText.length,\n          statusUpdated: statusResult.success,\n          timestamp\n        };\n\n      } catch (error) {\n        console.error(\"❌ Error processing blob extraction:\", error);\n        return {\n          success: false,\n          pageId,\n          timestamp,\n          error: \"Internal server error\",\n          details: error.message\n        };\n      }\n    }\n\n## Step 7: Add Helper Functions to Controller\n\nIn `/backend/controllers/tasks.controller.ts`, add these helper functions:\n\n    /**\n     * Extract PDF URL from Notion page properties\n     */\n    function extractPdfUrl(page: NotionPageProperties): string | null {\n      // Extract PDF property\n      const pdfProperty = page.properties?.PDF;\n      if (!pdfProperty || pdfProperty.type !== 'files') {\n        console.error(\"❌ No PDF property found or property is not of type 'files'\");\n        return null;\n      }\n\n      const files = pdfProperty.files;\n      if (!files || files.length === 0) {\n        console.error(\"❌ No files found in PDF property\");\n        return null;\n      }\n\n      // Get the first (and should be only) PDF file\n      const pdfFile = files[0];\n      if (!pdfFile) {\n        console.error(\"❌ PDF file is null or undefined\");\n        return null;\n      }\n\n      // Get the file URL (handle both external and Notion-hosted files)\n      if (pdfFile.type === 'external') {\n        return pdfFile.external?.url || null;\n      } else if (pdfFile.type === 'file') {\n        return pdfFile.file?.url || null;\n      } else {\n        console.error(\"❌ Unknown PDF file type:\", pdfFile.type);\n        return null;\n      }\n    }\n\n    /**\n     * Extract page ID from Notion webhook payload\n     */\n    export function extractPageIdFromWebhook(webhookBody: WebhookPayload): string | null {\n      const pageId = webhookBody.data?.id;\n      if (!pageId) {\n        console.error(\"❌ No page ID found in webhook payload\", JSON.stringify(webhookBody, null, 2));\n        return null;\n      }\n      return pageId;\n    }\n\n## Step 8: Verify Required Service Functions\n\nCheck that `/backend/services/notion.service.ts` contains these functions. If\nthey don't exist, add them:\n\n    import { Client } from \"npm:@notionhq/client\";\n\n    const notion = new Client({ auth: Deno.env.get(\"NOTION_API_KEY\") });\n\n    export async function getPageProperties(pageId: string) {\n      try {\n        const response = await notion.pages.retrieve({\n          page_id: pageId,\n        });\n        return {\n          success: true,\n          data: response,\n          timestamp: new Date().toISOString(),\n        };\n      } catch (error) {\n        return {\n          success: false,\n          error: error.message,\n          timestamp: new Date().toISOString(),\n        };\n      }\n    }\n\n    export async function updatePageStatus(pageId: string, status: string, blobKey?: string) {\n      try {\n        const properties: any = {\n          Status: {\n            select: {\n              name: status\n            }\n          }\n        };\n\n        // Add blob key if provided\n        if (blobKey) {\n          properties[\"Blob Key\"] = {\n            rich_text: [\n              {\n                text: {\n                  content: blobKey\n                }\n              }\n            ]\n          };\n        }\n\n        const response = await notion.pages.update({\n          page_id: pageId,\n          properties: properties\n        });\n\n        return {\n          success: true,\n          data: response,\n          timestamp: new Date().toISOString(),\n        };\n      } catch (error) {\n        return {\n          success: false,\n          error: error.message,\n          timestamp: new Date().toISOString(),\n        };\n      }\n    }\n\n## Step 9: Verify Required Utility Function\n\nCheck that `/backend/utils/pdfExtractor.ts` contains the `extractTextFromPDFUrl`\nfunction. If it doesn't exist, add it:\n\n    /**\n     * Extract text from a PDF file accessible via URL\n     */\n    export async function extractTextFromPDFUrl(pdfUrl: string): Promise {\n      try {\n        console.log(`📄 Downloading PDF from URL: ${pdfUrl}`);\n        \n        // Download the PDF file\n        const response = await fetch(pdfUrl);\n        if (!response.ok) {\n          throw new Error(`Failed to download PDF: ${response.status} ${response.statusText}`);\n        }\n        \n        const arrayBuffer = await response.arrayBuffer();\n        console.log(`📄 Downloaded PDF, size: ${arrayBuffer.byteLength} bytes`);\n        \n        // Use pdfjs-dist to extract text\n        const pdfjsLib = await import(\"https://esm.sh/pdfjs-dist@4.0.379/legacy/build/pdf.mjs\");\n        \n        // Load the PDF document\n        const loadingTask = pdfjsLib.getDocument({ data: arrayBuffer });\n        const pdfDocument = await loadingTask.promise;\n        \n        console.log(`📄 PDF loaded with ${pdfDocument.numPages} pages`);\n        \n        // Extract text from all pages\n        let fullText = '';\n        for (let pageNum = 1; pageNum <= pdfDocument.numPages; pageNum++) {\n          const page = await pdfDocument.getPage(pageNum);\n          const textContent = await page.getTextContent();\n          \n          // Combine text items into a single string\n          const pageText = textContent.items\n            .map((item: any) => item.str)\n            .join(' ');\n          \n          fullText += pageText + '\\n';\n          console.log(`📄 Extracted text from page ${pageNum}: ${pageText.length} characters`);\n        }\n        \n        console.log(`✅ Total extracted text: ${fullText.length} characters`);\n        return fullText.trim();\n        \n      } catch (error) {\n        console.error(\"❌ Error extracting text from PDF URL:\", error);\n        throw new Error(`PDF text extraction failed: ${error.message}`);\n      }\n    }\n\n## Step 10: Test the Implementation\n\n### Test 1: Basic Functionality Test\n\n    # Use the fetch tool to test the endpoint\n    fetch(\"/tasks/blob\", {\n      method: \"POST\",\n      body: JSON.stringify({\n        data: {\n          id: \"test-page-id-123\"\n        }\n      }),\n      headers: {\n        \"Content-Type\": \"application/json\"\n      }\n    })\n\n### Test 2: Error Handling Test\n\n    # Test with invalid payload\n    fetch(\"/tasks/blob\", {\n      method: \"POST\", \n      body: JSON.stringify({\n        data: {} // Missing id\n      }),\n      headers: {\n        \"Content-Type\": \"application/json\"\n      }\n    })\n\n### Test 3: Check Logs\n\n    # Use the requests tool to examine execution logs\n    requests(\"main.tsx\")\n\n## Step 11: Verification Checklist\n\nAfter implementation, verify:\n\n- \\[ \\] **Types file created**: `/backend/types/tasks.types.ts` exists with all\n  required interfaces\n- \\[ \\] **Route handler added**: Endpoint is in\n  `/backend/routes/tasks/_tasks.routes.ts` before catch-all handler\n- \\[ \\] **Imports correct**: All files have proper import statements with\n  correct paths\n- \\[ \\] **Controller function**: `processBlobExtraction` exists in tasks\n  controller with proper typing\n- \\[ \\] **Helper functions**: `extractPageIdFromWebhook` and `extractPdfUrl`\n  exist with proper typing\n- \\[ \\] **Service functions**: `getPageProperties` and `updatePageStatus` exist\n  in notion service\n- \\[ \\] **Utility function**: `extractTextFromPDFUrl` exists in PDF extractor\n  utility\n- \\[ \\] **Available endpoints updated**: New endpoint listed in catch-all\n  handler\n- \\[ \\] **Blob storage format**: Uses key format\n  `findings--transcripts--{pageId}`\n- \\[ \\] **Error handling**: Proper HTTP status codes (400 for client errors, 500\n  for server errors)\n- \\[ \\] **Logging consistency**: All console.log statements use emoji prefixes\n- \\[ \\] **Type safety**: No `any` types used except where necessary for external\n  APIs\n\n## Expected Behavior\n\n**Successful Request:**\n\n- Status: 200\n- Response:\n\n  { \"success\": true, \"pageId\": \"abc123\", \"blobKey\":\n  \"findings--transcripts--abc123\", \"textLength\": 1234, \"statusUpdated\": true,\n  \"timestamp\": \"2024-01-01T12:00:00.000Z\" }\n\n**Missing Page ID:**\n\n- Status: 400\n- Response:\n\n  { \"error\": \"Page ID is required\" }\n\n**No PDF Found:**\n\n- Status: 400\n- Response:\n\n  { \"success\": false, \"pageId\": \"abc123\", \"error\": \"No PDF files found in PDF\n  property\", \"timestamp\": \"2024-01-01T12:00:00.000Z\" }\n\n**Internal Error:**\n\n- Status: 500\n- Response:\n\n  { \"error\": \"Internal server error\", \"details\": \"Specific error message\",\n  \"timestamp\": \"2024-01-01T12:00:00.000Z\" }\n\n## Architecture Compliance\n\nThis implementation follows the established patterns:\n\n- **Routes**: Handle HTTP request/response only, delegate to controllers\n- **Controllers**: Orchestrate business logic, call services and utilities\n- **Services**: Make pure API calls to external systems (Notion)\n- **Utils**: Provide reusable utility functions (PDF processing)\n- **Types**: Centralized type definitions shared across modules\n- **Error Handling**: Consistent success/error response patterns\n- **Logging**: Emoji-prefixed console logs for easy debugging\n\nThe endpoint integrates seamlessly with existing webhook authentication\nmiddleware and follows the same patterns as other endpoints in the project.",[],[154,225],[154],{"_102":234,"_95":235,"_155":236,"_157":-5,"_158":137,"_159":160,"_161":162,"_104":237,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_169":238,"_98":595},"d0f9e7f4-7951-11f0-8810-0224a6c84d84","backend","0198aa63-0ee2-725e-8c9f-fc915e8caa81",["D",1755205275298],[239,292,317,341,446,486,525,588],{"_102":240,"_95":241,"_155":242,"_157":236,"_158":137,"_159":160,"_161":162,"_104":243,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_169":244,"_98":291},"d0f9b68a-7951-11f0-8810-0224a6c84d84","controllers","0198aa63-0ee2-725e-8c9f-952287ab9a39",["D",1755205275298],[245,255,264,273,282],{"_102":246,"_95":247,"_155":248,"_157":242,"_158":137,"_159":249,"_161":162,"_104":250,"_99":251,"_164":177,"_166":252,"_167":22,"_168":-5,"_169":253,"_98":254},"72e01762-8426-11f0-ad01-0224a6c84d84","README.md","0198aa63-0ee2-725e-8c9f-8c3d44adcdc9",39,["D",1756396111921],7,"# Controllers\n\nBusiness logic coordination between routes and services. Controllers should:\n\n- Handle environment variables and configuration\n- Coordinate multiple service calls\n- Implement business rules and validation\n- Handle service business logic layer (result.success)\n\n## Available Controllers\n\n- `emails.controller.ts` - Handles email PDF attachment processing workflow\n  - **PDF Processing**: Processes each PDF attachment as separate database page\n  - **Filename Parsing**: Extracts interviewee names from filenames using parentheses pattern\n  - **Status Tracking**: Manages upload status progression (\"Uploading...\" → \"Uploaded\" → \"Upload error\")\n  - **3-Step Upload**: Orchestrates Notion's file upload process (create → upload → attach)\n- `tasks.controller.ts` - Handles PDF extraction, blob storage, and page duplication workflows for Notion webhooks\n  - **PDF Text Extraction & Redaction**: Extracts text from PDFs and redacts person names (from Notion Name property) with \"Respondent\" before blob storage\n  - **Page Duplication**: Includes content filtering and H3-to-paragraph transformation for \"Key findings\" and \"Notable disagreements\"\n  - **Filtered Database Duplication**: Duplicates pages with child databases while filtering out specific properties (Confidence Score, Consensus Level, Evidence Summary, Methodology)\n- `health.controller.ts` - System health checks\n- `findings.controller.ts` - Handles findings analysis workflow using ChatGPT and Notion integration\n  - **Question Tracking**: Records \"Total questions\" count during initialization and \"Completed questions\" count upon finalization\n  - **Progress Monitoring**: Enables completion rate analysis in Notion dashboards",[],[235,241,247],{"_102":256,"_95":257,"_155":258,"_157":242,"_158":137,"_159":218,"_161":162,"_104":259,"_99":32,"_164":260,"_166":261,"_167":22,"_168":-5,"_169":262,"_98":263},"fdca5ffa-841b-11f0-bd95-0224a6c84d84","emails.controller.ts","fdca6086-841b-11f0-bd95-0224a6c84d84",["D",1756391620516],"script","/**\n * Emails Controller\n * Pure business logic for processing PDF attachments to Notion\n * \n * Responsibilities:\n * - Filter and validate PDF attachments\n * - Create Notion pages for each PDF\n * - Upload files using Notion's 3-step process\n * - Update page statuses based on results\n * - Return structured processing results\n * \n * Does NOT handle:\n * - Email communication (handled by email.service.ts)\n * - User notifications (handled by email handler orchestration)\n */\n\nimport {\n  createPageInDatabase,\n  createNotionFileUpload,\n  uploadFileToNotionUrl,\n  attachFileUploadToPage,\n  updatePageProperties,\n} from \"../services/notion.service.ts\";\nimport { extractIntervieweeName } from \"../utils/email.utils.ts\";\nimport type {\n  ProcessEmailAttachmentsRequest,\n  ProcessEmailAttachmentsResponse,\n  AttachmentProcessingResult,\n  NotionFileUploadObject,\n} from \"../types/email.types.ts\";\n\n/**\n * Main function to process email PDF attachments\n * Creates child pages in Notion database and uploads PDFs using 3-step process\n */\nexport async function processEmailPdfAttachments(\n  request: ProcessEmailAttachmentsRequest\n): Promise {\n  const timestamp = new Date().toISOString();\n  const results: AttachmentProcessingResult[] = [];\n  const errors: string[] = [];\n  const pdfPropertyName = request.pdfPropertyName || \"PDF\";\n  const batchId = `batch-${Date.now()}-${Math.random().toString(36).substr(2, 5)}`;\n\n  console.log(`📧 [${batchId}] Processing ${request.attachments.length} email attachments`);\n\n  // Filter for PDF attachments only using File object properties\n  const pdfAttachments = request.attachments.filter(\n    file => file.name.toLowerCase().endsWith('.pdf') || \n            file.type.includes('pdf')\n  );\n\n  console.log(`📄 [${batchId}] Found ${pdfAttachments.length} PDF attachments`);\n\n  if (pdfAttachments.length === 0) {\n    return {\n      success: true,\n      totalAttachments: request.attachments.length,\n      processedAttachments: 0,\n      successfulUploads: 0,\n      results: [],\n      errors: [\"No PDF attachments found\"],\n      timestamp,\n    };\n  }\n\n  // Phase 1: Create all Notion pages first (fast)\n  console.log(`📝 [${batchId}] Phase 1: Creating ${pdfAttachments.length} Notion pages...`);\n  const pageCreationResults: Array<{file: File, pageId?: string, error?: string}> = [];\n  \n  for (let i = 0; i < pdfAttachments.length; i++) {\n    const file = pdfAttachments[i];\n    console.log(`📄 [${batchId}] Creating page ${i + 1}/${pdfAttachments.length}: ${file.name}`);\n    \n    const pageResult = await createNotionPageForFile(file, request.databaseId);\n    pageCreationResults.push({\n      file,\n      pageId: pageResult.pageId,\n      error: pageResult.error\n    });\n    \n    if (pageResult.error) {\n      errors.push(`${file.name}: Page creation failed - ${pageResult.error}`);\n    }\n  }\n  \n  const successfulPages = pageCreationResults.filter(r => r.pageId);\n  console.log(`✅ [${batchId}] Phase 1 complete: ${successfulPages.length}/${pdfAttachments.length} pages created`);\n\n  // Phase 2: Upload files to created pages (slow)\n  console.log(`📤 [${batchId}] Phase 2: Uploading files to ${successfulPages.length} pages...`);\n  \n  for (let i = 0; i < pageCreationResults.length; i++) {\n    const { file, pageId, error: pageError } = pageCreationResults[i];\n    \n    const result: AttachmentProcessingResult = {\n      filename: file.name,\n      pageCreated: !!pageId,\n      fileUploaded: false,\n      pageId,\n    };\n    \n    if (pageError) {\n      result.error = pageError;\n      results.push(result);\n      continue;\n    }\n    \n    if (!pageId) {\n      result.error = \"No page ID available\";\n      results.push(result);\n      continue;\n    }\n    \n    console.log(`📤 [${batchId}] Uploading file ${i + 1}/${pageCreationResults.length}: ${file.name}`);\n    \n    // Upload file to the created page\n    const uploadResult = await uploadPdfToNotionPage(pageId, file, pdfPropertyName);\n    \n    if (uploadResult.success) {\n      result.fileUploaded = true;\n      result.uploadId = uploadResult.uploadId;\n      console.log(`✅ [${batchId}] File uploaded successfully: ${file.name}`);\n      \n      // Update status to \"Uploaded\"\n      await updatePageProperties(pageId, {\n        \"Status\": { select: { name: \"Uploaded\" } }\n      });\n      \n    } else {\n      result.error = `File upload failed: ${uploadResult.error}`;\n      console.log(`❌ [${batchId}] File upload failed for ${file.name}: ${uploadResult.error}`);\n      \n      // Update status to \"Upload error\"\n      await updatePageProperties(pageId, {\n        \"Status\": { select: { name: \"Upload error\" } }\n      });\n      \n      errors.push(`${file.name}: ${uploadResult.error}`);\n    }\n    \n    results.push(result);\n    \n    // Small delay between uploads to avoid rate limiting\n    if (i < pageCreationResults.length - 1) {\n      await new Promise(resolve => setTimeout(resolve, 500));\n    }\n  }\n\n  const successfulUploads = results.filter(r => r.fileUploaded).length;\n  const processedAttachments = results.filter(r => r.pageCreated).length;\n\n  console.log(`📊 [${batchId}] Processing complete: ${processedAttachments}/${pdfAttachments.length} pages created, ${successfulUploads}/${pdfAttachments.length} files uploaded`);\n\n  return {\n    success: errors.length === 0,\n    totalAttachments: request.attachments.length,\n    processedAttachments,\n    successfulUploads,\n    results,\n    errors,\n    timestamp,\n  };\n}\n\n/**\n * Create a Notion page for a PDF file (Phase 1 - fast)\n */\nasync function createNotionPageForFile(\n  file: File,\n  databaseId: string\n): Promise<{ pageId?: string; error?: string }> {\n  try {\n    // Extract interviewee name from filename\n    const intervieweeName = extractIntervieweeName(file.name);\n    console.log(`   - Extracted interviewee name: \"${intervieweeName}\"`);\n\n    // Create page properties\n    const pageProperties = {\n      \"Name\": {\n        title: [\n          {\n            text: {\n              content: intervieweeName\n            }\n          }\n        ]\n      },\n      \"Filename\": {\n        rich_text: [\n          {\n            text: {\n              content: file.name\n            }\n          }\n        ]\n      },\n      \"Status\": {\n        select: {\n          name: \"Uploading...\"\n        }\n      }\n    };\n\n    const pageResult = await createPageInDatabase(databaseId, pageProperties);\n    \n    if (!pageResult.success) {\n      return { error: `Failed to create page: ${pageResult.error}` };\n    }\n\n    console.log(`✅ Page created: ${pageResult.data.id}`);\n    return { pageId: pageResult.data.id };\n\n  } catch (error) {\n    console.error(`❌ Error creating page for ${file.name}:`, error);\n    return { error: `Unexpected error: ${error.message}` };\n  }\n}\n\n/**\n * Upload PDF to Notion page using exact 3-step file upload process\n */\nasync function uploadPdfToNotionPage(\n  pageId: string,\n  pdfFile: File,\n  propertyName: string\n): Promise<{ success: boolean; uploadId?: string; error?: string }> {\n  try {\n    // Validate file before upload\n    console.log(`📋 Validating file: ${pdfFile.name}`);\n    console.log(`   - Content type: ${pdfFile.type}`);\n    console.log(`   - File size: ${pdfFile.size} bytes`);\n    \n    if (pdfFile.size === 0) {\n      return { success: false, error: 'File is empty' };\n    }\n    \n    if (pdfFile.size > 50 * 1024 * 1024) { // 50MB limit\n      return { success: false, error: 'File too large (>50MB)' };\n    }\n\n    // Step 1: Create File Upload Object\n    console.log(`📤 Step 1: Creating upload object for ${pdfFile.name}`);\n    const createUploadResult = await createNotionFileUpload();\n    \n    if (!createUploadResult.success) {\n      return { success: false, error: `Step 1 failed: ${createUploadResult.error}` };\n    }\n\n    const uploadObject = createUploadResult.data;\n    console.log(`✅ Step 1 complete: Upload object created with ID ${uploadObject.id}`);\n\n    // Step 2: Upload File Contents\n    console.log(`📤 Step 2: Uploading file contents for ${pdfFile.name}`);\n    \n    const uploadResult = await uploadFileToNotionUrl(uploadObject.upload_url, pdfFile);\n    \n    if (!uploadResult.success) {\n      return { success: false, error: `Step 2 failed: ${uploadResult.error}` };\n    }\n\n    console.log(`✅ Step 2 complete: File contents uploaded`);\n    console.log(`   - Upload status: ${uploadResult.data.status}`);\n\n    // Step 3: Attach File to Page\n    console.log(`📤 Step 3: Attaching file to page ${pageId}`);\n    const attachResult = await attachFileUploadToPage(\n      pageId,\n      propertyName,\n      uploadObject.id,  // Use the upload object ID from Step 1\n      pdfFile.name\n    );\n\n    if (!attachResult.success) {\n      return { success: false, error: `Step 3 failed: ${attachResult.error}` };\n    }\n\n    console.log(`✅ Step 3 complete: File attached to page with type \"file_upload\"`);\n    console.log(`   - Property updated: ${propertyName}`);\n\n    return { success: true, uploadId: uploadObject.id };\n\n  } catch (error) {\n    console.error(`❌ Upload process error:`, error);\n    return { success: false, error: `Upload process failed: ${error.message}` };\n  }\n}",[],[235,241,257],{"_102":265,"_95":266,"_155":267,"_157":242,"_158":137,"_159":268,"_161":162,"_104":269,"_99":209,"_164":260,"_166":270,"_167":22,"_168":-5,"_169":271,"_98":272},"8ab4cbfa-840b-11f0-a09a-0224a6c84d84","findings.controller.ts","82c445de-7a05-11f0-9236-0224a6c84d84",37,["D",1756384555490],"/**\n * Findings Controller\n * Business logic for findings analysis workflow\n */\n\nimport { blob } from \"https://esm.town/v/std/blob\";\nimport {\n  queryDatabasePages,\n  getDatabasePages,\n  updatePageContent,\n  updatePageStatus,\n  deletePageContent,\n  appendPageContent,\n  updatePageProperties,\n  createInlineDatabase,\n  createPageInDatabaseWithContent,\n  getPageProperties,\n} from \"../services/notion.service.ts\";\nimport {\n  analyzeTranscriptsForQuestion,\n  analyzeTranscriptsWithConsensus,\n} from \"../services/openai.service.ts\";\nimport { createTimestampProperties, createAnalysisResultsDatabaseSchema, createResultsDatabaseProperty, createTotalQuestionsProperty, createCompletedQuestionsProperty, createEvidenceSummaryProperty, createMethodologyProperty } from \"../utils/notion-properties.utils.ts\";\nimport type { NotionProperties } from \"../types/notion.types.ts\";\nimport {\n  calculateDuration,\n  createTimingSummaryBlocks,\n} from \"../utils/process-timing.utils.ts\";\nimport type {\n  ProcessFindingsRequest,\n  ProcessFindingsResponse,\n  FindingsQuestion,\n  TranscriptPage,\n  TranscriptBlob,\n  QuestionAnalysis,\n  EnhancedQuestionAnalysis,\n  ConsensusAnswer,\n} from \"../types/findings.types.ts\";\n\n/**\n * Main function to process findings analysis with batched processing\n */\nexport async function processFindings(\n  request: ProcessFindingsRequest\n): Promise {\n  const timestamp = new Date().toISOString();\n  let reportPageId: string | undefined;\n  let resultsDatabaseId: string | undefined;\n  const BATCH_SIZE = 5; // Process 5 questions per batch\n\n  try {\n    console.log(\"🔍 Starting findings analysis process...\");\n\n    // Get environment variables\n    const reportsDbId = Deno.env.get(\"FINDINGS_REPORTS_RAW_DB_ID\");\n    const questionsDbId = Deno.env.get(\"FINDINGS_QUESTIONS_DB_ID\");\n    const transcriptsDbId = Deno.env.get(\"FINDINGS_TRANSCRIPTS_DB_ID\");\n    const openaiApiKey = Deno.env.get(\"OPENAI_API_KEY\");\n\n    if (!reportsDbId || !questionsDbId || !transcriptsDbId || !openaiApiKey) {\n      return {\n        success: false,\n        error: \"Missing required environment variables\",\n        timestamp,\n      };\n    }\n\n    // Step 1: Get report page (Status = \"Ready\" OR \"Working...\")\n    console.log(\"📊 Querying reports database for 'Ready' status...\");\n    const notStartedReportsResult = await queryDatabasePages(reportsDbId, {\n      property: \"Status\",\n      select: {\n        equals: \"Ready\",\n      },\n    });\n\n    console.log(\"📊 Querying reports database for 'Working...' status...\");\n    const workingReportsResult = await queryDatabasePages(reportsDbId, {\n      property: \"Status\",\n      select: {\n        equals: \"Working...\",\n      },\n    });\n\n    if (!notStartedReportsResult.success || !workingReportsResult.success) {\n      return {\n        success: false,\n        error: `Failed to query reports database: ${\n          notStartedReportsResult.error || workingReportsResult.error\n        }`,\n        timestamp,\n      };\n    }\n\n    const notStartedReports = notStartedReportsResult.data.results;\n    const workingReports = workingReportsResult.data.results;\n\n    if (notStartedReports.length === 0 && workingReports.length === 0) {\n      return {\n        success: false,\n        error:\n          \"No report pages found with Status 'Ready' or 'Working...'\",\n        timestamp,\n      };\n    }\n\n    // Prioritize \"Ready\" reports over \"Working...\" reports\n    if (notStartedReports.length > 0) {\n      // New report found - initialize it\n      reportPageId = notStartedReports[0].id;\n      console.log(`📄 Found new report page: ${reportPageId}`);\n\n      // Initialize: Reset all questions to \"Not started\" and create results database\n      resultsDatabaseId = await initializeNewReport(reportPageId, questionsDbId);\n      if (!resultsDatabaseId) {\n        return {\n          success: false,\n          error: \"Failed to initialize new report and create results database\",\n          timestamp,\n        };\n      }\n    } else {\n      // Continue with existing \"Working...\" report\n      reportPageId = workingReports[0].id;\n      console.log(`📄 Continuing with working report page: ${reportPageId}`);\n      \n      // For existing reports, we need to find the existing results database\n      // This is a simplified approach - in production you might want to store the database ID\n      console.log(\"🔍 Looking for existing Analysis Results database...\");\n      // For now, we'll create a new database if we can't find one\n      // In a more robust implementation, you'd store the database ID in the report page properties\n      resultsDatabaseId = await findOrCreateResultsDatabase(reportPageId);\n      if (!resultsDatabaseId) {\n        return {\n          success: false,\n          error: \"Failed to find or create results database for existing report\",\n          timestamp,\n        };\n      }\n    }\n\n    // Step 2: Get next batch of questions to process\n    console.log(\n      `❓ Fetching next batch of ${BATCH_SIZE} questions with 'Not started' status...`\n    );\n    const questionsResult = await queryDatabasePages(questionsDbId, {\n      property: \"Status\",\n      select: {\n        equals: \"Not started\",\n      },\n    });\n\n    if (!questionsResult.success) {\n      return {\n        success: false,\n        error: `Failed to get questions: ${questionsResult.error}`,\n        timestamp,\n      };\n    }\n\n    const allNotStartedQuestions = extractQuestions(\n      questionsResult.data.results\n    );\n\n    if (allNotStartedQuestions.length === 0) {\n      // No more questions to process - check if we're done\n      return await checkAndFinalizeReport(\n        reportPageId,\n        questionsDbId,\n        timestamp\n      );\n    }\n\n    // Take only the batch size we can handle\n    const questionsToProcess = allNotStartedQuestions.slice(0, BATCH_SIZE);\n    console.log(\n      `📝 Processing batch of ${questionsToProcess.length} questions (${allNotStartedQuestions.length} total remaining)`\n    );\n\n    // Step 3: Mark selected questions as \"Working...\"\n    console.log(\"🔄 Marking selected questions as 'Working...'\");\n    for (const question of questionsToProcess) {\n      const updateResult = await updatePageProperties(question.id, {\n        Status: {\n          select: {\n            name: \"Working...\",\n          },\n        },\n      });\n\n      if (!updateResult.success) {\n        console.error(\n          `⚠️ Failed to update question ${question.id} status: ${updateResult.error}`\n        );\n      }\n    }\n\n    // Step 4: Get transcript data (only once per batch)\n    console.log(\"📋 Fetching transcript pages...\");\n    const transcriptsResult = await getDatabasePages(transcriptsDbId);\n    if (!transcriptsResult.success) {\n      return {\n        success: false,\n        error: `Failed to get transcripts: ${transcriptsResult.error}`,\n        timestamp,\n      };\n    }\n\n    const transcriptPages = extractTranscriptPages(\n      transcriptsResult.data.results\n    );\n    console.log(`🗂️ Found ${transcriptPages.length} transcript pages`);\n\n    // Step 5: Fetch transcript blobs\n    console.log(\"💾 Fetching transcript blobs...\");\n    const transcriptTexts = await fetchTranscriptBlobs(\n      transcriptPages.map((t) => t.blobKey)\n    );\n    console.log(\n      `📄 Successfully fetched ${transcriptTexts.length} transcripts`\n    );\n\n    // Step 6: Process each question in the batch\n    console.log(\"🤖 Processing questions batch...\");\n    const batchStartTime = new Date();\n\n    let questionsProcessed = 0;\n    let questionsSucceeded = 0;\n\n    for (const question of questionsToProcess) {\n      console.log(\n        `🔍 Analyzing question ${questionsProcessed + 1}/${\n          questionsToProcess.length\n        }: ${question.name}`\n      );\n\n      try {\n        const analysisResult = await analyzeTranscriptsWithConsensus({\n          question: question.name,\n          transcripts: transcriptTexts,\n          apiKey: openaiApiKey,\n        });\n\n        if (analysisResult.success && analysisResult.data) {\n          // Create child page in results database\n          const consensus = analysisResult.data;\n          \n          // Create evidence summary text\n          const evidenceSummaryText = createEvidenceSummaryText(consensus.perTranscriptFindings);\n          \n          // Prepare page properties\n          const pageProperties = {\n            \"Name\": {\n              title: [\n                {\n                  text: {\n                    content: question.name\n                  }\n                }\n              ]\n            },\n            \"Consensus Level\": {\n              select: {\n                name: consensus.consensusLevel === \"high\" ? \"High\" : \n                       consensus.consensusLevel === \"medium\" ? \"Medium\" : \"Low\"\n              }\n            },\n            \"Confidence Score\": {\n              number: consensus.confidence\n            },\n            \"Status\": {\n              select: {\n                name: \"Done\"\n              }\n            },\n            \"Evidence Summary\": createEvidenceSummaryProperty(evidenceSummaryText),\n            \"Methodology\": createMethodologyProperty(consensus.methodologyNote || \"\")\n          };\n\n          // Prepare page content blocks\n          const contentBlocks = formatEnhancedReportContent([{\n            question: question.name,\n            consensusAnswer: consensus\n          }]);\n\n          // Create the child page with content\n          const createPageResult = await createPageInDatabaseWithContent(\n            resultsDatabaseId!,\n            pageProperties,\n            contentBlocks\n          );\n\n          if (createPageResult.success) {\n            questionsSucceeded++;\n            console.log(\n              `✅ Question ${\n                questionsProcessed + 1\n              } completed and added as child page (Consensus: ${\n                consensus.consensusLevel\n              }, Confidence: ${Math.round(\n                consensus.confidence * 100\n              )}%)`\n            );\n\n            // Mark question as \"Done\"\n            await updatePageProperties(question.id, {\n              Status: {\n                select: {\n                  name: \"Done\",\n                },\n              },\n            });\n          } else {\n            console.error(\n              `❌ Failed to create child page for question ${\n                questionsProcessed + 1\n              }: ${createPageResult.error}`\n            );\n            \n            // Create error page\n            await createErrorPage(\n              resultsDatabaseId!,\n              question.name,\n              `Failed to create result page: ${createPageResult.error}`\n            );\n\n            // Mark question as \"Error\"\n            await updatePageProperties(question.id, {\n              Status: {\n                select: {\n                  name: \"Error\",\n                },\n              },\n            });\n          }\n        } else {\n          console.error(\n            `❌ Failed to analyze question ${questionsProcessed + 1}: ${\n              analysisResult.error\n            }`\n          );\n          \n          // Create error page for failed analysis\n          await createErrorPage(\n            resultsDatabaseId!,\n            question.name,\n            `Analysis failed: ${analysisResult.error}`\n          );\n\n          // Mark question as \"Error\"\n          await updatePageProperties(question.id, {\n            Status: {\n              select: {\n                name: \"Error\",\n              },\n            },\n          });\n        }\n      } catch (error) {\n        console.error(\n          `❌ Unexpected error processing question ${questionsProcessed + 1}:`,\n          error\n        );\n        \n        // Create error page for unexpected errors\n        await createErrorPage(\n          resultsDatabaseId!,\n          question.name,\n          `Unexpected error: ${error.message}`\n        );\n\n        // Mark question as \"Error\"\n        await updatePageProperties(question.id, {\n          Status: {\n            select: {\n              name: \"Error\",\n            },\n          },\n        });\n      }\n\n      questionsProcessed++;\n    }\n\n    // Calculate batch processing time\n    const batchEndTime = new Date();\n    const batchDuration = calculateDuration(batchStartTime, batchEndTime);\n    console.log(\n      `⏱️ Batch processing time: ${batchDuration.minutes}m ${batchDuration.seconds}s`\n    );\n\n    // Check if all questions are now complete\n    console.log(\"🔍 Checking for remaining incomplete questions...\");\n    const incompleteQuestionsResult = await queryDatabasePages(questionsDbId, {\n      property: \"Status\",\n      select: {\n        does_not_equal: \"Done\",\n      },\n    });\n\n    const remainingQuestions = incompleteQuestionsResult.success\n      ? incompleteQuestionsResult.data.results.length\n      : 0;\n    console.log(`📊 Found ${remainingQuestions} questions that are not 'Done'`);\n\n    if (remainingQuestions === 0) {\n      // All questions processed - finalize the report\n      console.log(\"🎉 All questions processed! Finalizing report...\");\n      return await finalizeReport(reportPageId, questionsDbId, timestamp);\n    } else {\n      console.log(\n        `📊 Batch complete. ${remainingQuestions} questions remaining for next cron run.`\n      );\n\n      return {\n        success: true,\n        reportPageId,\n        questionsProcessed: questionsToProcess.length,\n        questionsSucceeded,\n        transcriptsAnalyzed: transcriptTexts.length,\n        timestamp,\n        remainingQuestions,\n      };\n    }\n  } catch (error) {\n    console.error(\"❌ Error in findings analysis:\", error);\n\n    // Try to record error status if we have a report page\n    if (reportPageId) {\n      const errorEndTime = new Date();\n      console.log(`⏱️ Recording error end time: ${errorEndTime.toISOString()}`);\n\n      const errorStatusResult = await updatePageProperties(\n        reportPageId,\n        createTimestampProperties(undefined, errorEndTime, \"Error\")\n      );\n\n      if (!errorStatusResult.success) {\n        console.error(\n          `⚠️ Failed to update error status and end time: ${errorStatusResult.error}`\n        );\n      } else {\n        console.log(\"✅ Error status and end time recorded in database\");\n      }\n    }\n\n    return {\n      success: false,\n      error: \"Internal server error\",\n      details: error.message,\n      timestamp,\n    };\n  }\n}\n\n/**\n * Initialize a new report by resetting all questions and creating the results database\n */\nasync function initializeNewReport(\n  reportPageId: string,\n  questionsDbId: string\n): Promise {\n  console.log(\"🔄 Initializing new report...\");\n\n  // Capture and record start time immediately\n  const processStartTime = new Date();\n  console.log(`⏱️ Recording start time: ${processStartTime.toISOString()}`);\n\n  const startTimeResult = await updatePageProperties(\n    reportPageId,\n    createTimestampProperties(processStartTime, undefined, \"Working...\")\n  );\n\n  if (!startTimeResult.success) {\n    console.error(`⚠️ Failed to update start time: ${startTimeResult.error}`);\n  } else {\n    console.log(\"✅ Start time recorded in database\");\n  }\n\n  // Clear page content\n  console.log(\"🗑️ Clearing existing page content...\");\n  const clearContentResult = await deletePageContent(reportPageId);\n  if (!clearContentResult.success) {\n    console.error(\n      `⚠️ Failed to clear page content: ${clearContentResult.error}`\n    );\n  } else {\n    console.log(\n      `🗑️ Cleared ${clearContentResult.data.deletedBlocks} existing blocks`\n    );\n  }\n\n  // Create inline database for analysis results\n  console.log(\"📊 Creating Analysis Results database...\");\n  const databaseProperties = createAnalysisResultsDatabaseSchema();\n\n  const createDbResult = await createInlineDatabase(\n    reportPageId,\n    \"Analysis Results\",\n    databaseProperties\n  );\n\n  if (!createDbResult.success) {\n    console.error(`❌ Failed to create inline database: ${createDbResult.error}`);\n    return null;\n  }\n\n  const resultsDatabaseId = createDbResult.data.databaseId;\n  console.log(`✅ Created Analysis Results database: ${resultsDatabaseId}`);\n\n  // Store the database ID in the report page properties\n  console.log(\"💾 Storing database ID in report page properties...\");\n  const storeDatabaseIdResult = await updatePageProperties(\n    reportPageId,\n    createResultsDatabaseProperty(resultsDatabaseId)\n  );\n\n  if (!storeDatabaseIdResult.success) {\n    console.error(`⚠️ Failed to store database ID: ${storeDatabaseIdResult.error}`);\n  } else {\n    console.log(\"✅ Database ID stored in report page properties\");\n  }\n\n  // Reset all questions to \"Not started\"\n  console.log(\"🔄 Resetting all questions to 'Not started' status...\");\n  const allQuestionsResult = await getDatabasePages(questionsDbId);\n\n  if (allQuestionsResult.success) {\n    const allQuestions = extractQuestions(allQuestionsResult.data.results);\n    console.log(\n      `📝 Resetting ${allQuestions.length} questions to 'Not started'`\n    );\n\n    // Update report page with total questions count\n    console.log(`📊 Recording total questions count: ${allQuestions.length}`);\n    const totalQuestionsResult = await updatePageProperties(\n      reportPageId,\n      {\n        \"Total questions\": createTotalQuestionsProperty(allQuestions.length)\n      }\n    );\n\n    if (!totalQuestionsResult.success) {\n      console.error(`⚠️ Failed to update total questions count: ${totalQuestionsResult.error}`);\n    } else {\n      console.log(\"✅ Total questions count recorded in database\");\n    }\n\n    for (const question of allQuestions) {\n      const updateResult = await updatePageProperties(question.id, {\n        Status: {\n          select: {\n            name: \"Not started\",\n          },\n        },\n      });\n\n      if (!updateResult.success) {\n        console.error(\n          `⚠️ Failed to reset question ${question.id} status: ${updateResult.error}`\n        );\n      }\n    }\n\n    console.log(\"✅ All questions reset to 'Not started'\");\n  } else {\n    console.error(\n      `⚠️ Failed to get questions for reset: ${allQuestionsResult.error}`\n    );\n  }\n\n  return resultsDatabaseId;\n}\n\n/**\n * Find existing results database or create a new one for continuing reports\n */\nasync function findOrCreateResultsDatabase(reportPageId: string): Promise {\n  try {\n    // First, try to get the database ID from report page properties\n    console.log(\"🔍 Looking for existing Results Database ID in page properties...\");\n    const pagePropertiesResult = await getPageProperties(reportPageId);\n    \n    if (pagePropertiesResult.success) {\n      const resultsDatabaseId = extractResultsDatabaseId(pagePropertiesResult.data);\n      if (resultsDatabaseId) {\n        console.log(`✅ Found existing Results Database ID: ${resultsDatabaseId}`);\n        return resultsDatabaseId;\n      }\n    }\n    \n    console.log(\"📊 No existing database ID found, creating new Analysis Results database...\");\n    const databaseProperties = createAnalysisResultsDatabaseSchema();\n\n    const createDbResult = await createInlineDatabase(\n      reportPageId,\n      \"Analysis Results\",\n      databaseProperties\n    );\n\n    if (!createDbResult.success) {\n      console.error(`❌ Failed to create results database: ${createDbResult.error}`);\n      return null;\n    }\n\n    const resultsDatabaseId = createDbResult.data.databaseId;\n    console.log(`✅ Created Analysis Results database: ${resultsDatabaseId}`);\n    \n    // Store the database ID in the report page properties\n    console.log(\"💾 Storing database ID in report page properties...\");\n    const storeDatabaseIdResult = await updatePageProperties(\n      reportPageId,\n      createResultsDatabaseProperty(resultsDatabaseId)\n    );\n\n    if (!storeDatabaseIdResult.success) {\n      console.error(`⚠️ Failed to store database ID: ${storeDatabaseIdResult.error}`);\n    } else {\n      console.log(\"✅ Database ID stored in report page properties\");\n    }\n    \n    return resultsDatabaseId;\n  } catch (error) {\n    console.error(`❌ Error finding/creating results database:`, error);\n    return null;\n  }\n}\n\n/**\n * Create an error page in the results database\n */\nasync function createErrorPage(\n  databaseId: string,\n  questionName: string,\n  errorMessage: string\n): Promise {\n  try {\n    const pageProperties = {\n      \"Name\": {\n        title: [\n          {\n            text: {\n              content: questionName\n            }\n          }\n        ]\n      },\n      \"Consensus Level\": {\n        select: {\n          name: \"Low\"\n        }\n      },\n      \"Confidence Score\": {\n        number: 0\n      },\n      \"Status\": {\n        select: {\n          name: \"Error\"\n        }\n      },\n      \"Evidence Summary\": createEvidenceSummaryProperty(\"\"),\n      \"Methodology\": createMethodologyProperty(\"\")\n    };\n\n    const errorContent = [\n      {\n        object: \"block\",\n        type: \"callout\",\n        callout: {\n          icon: { emoji: \"❌\" },\n          rich_text: [\n            {\n              type: \"text\",\n              text: {\n                content: `Error: ${errorMessage}`,\n              },\n              annotations: { bold: true, color: \"red\" },\n            },\n          ],\n        },\n      },\n    ];\n\n    const createResult = await createPageInDatabaseWithContent(\n      databaseId,\n      pageProperties,\n      errorContent\n    );\n\n    if (!createResult.success) {\n      console.error(`❌ Failed to create error page: ${createResult.error}`);\n    }\n  } catch (error) {\n    console.error(`❌ Error creating error page:`, error);\n  }\n}\nasync function checkAndFinalizeReport(\n  reportPageId: string,\n  questionsDbId: string,\n  timestamp: string\n): Promise {\n  console.log(\"🔍 Checking if all questions are complete...\");\n\n  // Check for any remaining incomplete questions (any status other than \"Done\")\n  const incompleteQuestionsResult = await queryDatabasePages(questionsDbId, {\n    property: \"Status\",\n    select: {\n      does_not_equal: \"Done\",\n    },\n  });\n\n  if (\n    incompleteQuestionsResult.success &&\n    incompleteQuestionsResult.data.results.length > 0\n  ) {\n    const incompleteCount = incompleteQuestionsResult.data.results.length;\n    console.log(\n      `⚠️ Found ${incompleteCount} questions that are not 'Done'. Report is not ready for finalization.`\n    );\n\n    // Log the statuses for debugging\n    const statusCounts: Record = {};\n    for (const question of incompleteQuestionsResult.data.results) {\n      const status = question.properties?.Status?.select?.name || \"Unknown\";\n      statusCounts[status] = (statusCounts[status] || 0) + 1;\n    }\n    console.log(`📊 Incomplete question statuses:`, statusCounts);\n\n    return {\n      success: false,\n      error: `Report not ready: ${incompleteCount} questions are not complete`,\n      timestamp,\n      remainingQuestions: incompleteCount,\n    };\n  }\n\n  return await finalizeReport(reportPageId, questionsDbId, timestamp);\n}\n\n/**\n * Finalize the report by updating status and properties (no content summary)\n */\nasync function finalizeReport(\n  reportPageId: string,\n  questionsDbId: string,\n  timestamp: string\n): Promise {\n  console.log(\"🎉 Finalizing report...\");\n\n  // Get final counts\n  const doneQuestionsResult = await queryDatabasePages(questionsDbId, {\n    property: \"Status\",\n    select: {\n      equals: \"Done\",\n    },\n  });\n\n  const errorQuestionsResult = await queryDatabasePages(questionsDbId, {\n    property: \"Status\",\n    select: {\n      equals: \"Error\",\n    },\n  });\n\n  const allQuestionsResult = await getDatabasePages(questionsDbId);\n\n  const questionsSucceeded = doneQuestionsResult.success\n    ? doneQuestionsResult.data.results.length\n    : 0;\n  const questionsErrored = errorQuestionsResult.success\n    ? errorQuestionsResult.data.results.length\n    : 0;\n  const totalQuestions = allQuestionsResult.success\n    ? allQuestionsResult.data.results.length\n    : 0;\n\n  console.log(\n    `📊 Final counts: ${questionsSucceeded} succeeded, ${questionsErrored} errors, ${totalQuestions} total`\n  );\n\n  // Update report page with completed questions count, end time, and final status in a single atomic call\n  const processEndTime = new Date();\n  console.log(\n    `📊 Recording completed questions count (${questionsSucceeded}), end time, and updating status to 'Done'...`\n  );\n  \n  const finalProperties: NotionProperties = {\n    \"Completed questions\": createCompletedQuestionsProperty(questionsSucceeded),\n    ...createTimestampProperties(undefined, processEndTime, \"Done\")\n  };\n  \n  const finalStatusResult = await updatePageProperties(\n    reportPageId,\n    finalProperties\n  );\n\n  if (!finalStatusResult.success) {\n    console.error(\n      `⚠️ Failed to update completed questions count, end time and status: ${finalStatusResult.error}`\n    );\n  } else {\n    console.log(\n      \"✅ Completed questions count recorded, end time updated, and report page status set to 'Done'\"\n    );\n  }\n\n  const successMessage =\n    questionsErrored === 0\n      ? \"✅ All questions processed successfully\"\n      : `⚠️ Completed with ${questionsErrored} errors: ${questionsSucceeded}/${totalQuestions} questions processed successfully`;\n\n  console.log(successMessage);\n\n  return {\n    success: true,\n    reportPageId,\n    questionsProcessed: totalQuestions,\n    questionsSucceeded,\n    transcriptsAnalyzed: 0, // Not tracked in batched mode\n    timestamp,\n  };\n}\n\n/**\n * Extract questions from Notion database pages\n */\nfunction extractQuestions(pages: any[]): FindingsQuestion[] {\n  return pages\n    .map((page) => ({\n      id: page.id,\n      name: extractNameProperty(page),\n    }))\n    .filter((q) => q.name); // Filter out pages without names\n}\n\n/**\n * Extract transcript pages with blob keys\n */\nfunction extractTranscriptPages(pages: any[]): TranscriptPage[] {\n  return pages\n    .map((page) => ({\n      id: page.id,\n      blobKey: extractBlobKeyProperty(page),\n    }))\n    .filter((t) => t.blobKey); // Filter out pages without blob keys\n}\n\n/**\n * Fetch transcript blobs and extract text\n */\nasync function fetchTranscriptBlobs(blobKeys: string[]): Promise {\n  const transcriptTexts: string[] = [];\n\n  for (const blobKey of blobKeys) {\n    try {\n      const blobData = (await blob.getJSON(blobKey)) as TranscriptBlob;\n      if (blobData && blobData.extractedText) {\n        transcriptTexts.push(blobData.extractedText);\n      }\n    } catch (error) {\n      console.error(`⚠️ Failed to fetch blob ${blobKey}:`, error);\n      // Continue with other blobs\n    }\n  }\n\n  return transcriptTexts;\n}\n\n/**\n * Format questions and answers into Notion content blocks (legacy)\n */\nfunction formatReportContent(analyses: QuestionAnalysis[]): any[] {\n  const blocks: any[] = [];\n\n  for (const analysis of analyses) {\n    // Add question as heading\n    blocks.push({\n      object: \"block\",\n      type: \"heading_2\",\n      heading_2: {\n        rich_text: [\n          {\n            type: \"text\",\n            text: {\n              content: analysis.question,\n            },\n          },\n        ],\n      },\n    });\n\n    // Add answer as paragraph\n    blocks.push({\n      object: \"block\",\n      type: \"paragraph\",\n      paragraph: {\n        rich_text: [\n          {\n            type: \"text\",\n            text: {\n              content: analysis.answer,\n            },\n          },\n        ],\n      },\n    });\n\n    // Add spacing\n    blocks.push({\n      object: \"block\",\n      type: \"paragraph\",\n      paragraph: {\n        rich_text: [],\n      },\n    });\n  }\n\n  return blocks;\n}\n\n/**\n * Format enhanced consensus analysis into rich Notion content blocks\n * Note: Evidence Summary and Methodology are now stored as database properties\n */\nfunction formatEnhancedReportContent(\n  analyses: EnhancedQuestionAnalysis[]\n): any[] {\n  const blocks: any[] = [];\n\n  for (const analysis of analyses) {\n    const consensus = analysis.consensusAnswer;\n\n    // Add direct answer (no question heading since it's in the Name property)\n    blocks.push({\n      object: \"block\",\n      type: \"paragraph\",\n      paragraph: {\n        rich_text: [\n          {\n            type: \"text\",\n            text: {\n              content: consensus.directAnswer,\n            },\n          },\n        ],\n      },\n    });\n\n    // Add key points if any (h3 heading like Notable Disagreements)\n    if (consensus.keyPoints.length > 0) {\n      blocks.push({\n        object: \"block\",\n        type: \"heading_3\",\n        heading_3: {\n          rich_text: [\n            {\n              type: \"text\",\n              text: {\n                content: \"Key Findings\",\n              },\n            },\n          ],\n        },\n      });\n\n      for (const point of consensus.keyPoints) {\n        blocks.push({\n          object: \"block\",\n          type: \"bulleted_list_item\",\n          bulleted_list_item: {\n            rich_text: [\n              {\n                type: \"text\",\n                text: {\n                  content: point,\n                },\n              },\n            ],\n          },\n        });\n      }\n    }\n\n    // Add disagreements if any\n    if (consensus.disagreements.length > 0) {\n      blocks.push({\n        object: \"block\",\n        type: \"heading_3\",\n        heading_3: {\n          rich_text: [\n            {\n              type: \"text\",\n              text: {\n                content: \"Notable Disagreements\",\n              },\n            },\n          ],\n        },\n      });\n\n      for (const disagreement of consensus.disagreements) {\n        blocks.push({\n          object: \"block\",\n          type: \"bulleted_list_item\",\n          bulleted_list_item: {\n            rich_text: [\n              {\n                type: \"text\",\n                text: {\n                  content: disagreement,\n                },\n              },\n            ],\n          },\n        });\n      }\n    }\n\n    // Evidence Summary and Methodology are now stored as database properties\n    // They are no longer included in the page content\n\n    // Add spacing between questions (only needed if processing multiple questions in one page)\n    blocks.push({\n      object: \"block\",\n      type: \"divider\",\n      divider: {},\n    });\n\n    blocks.push({\n      object: \"block\",\n      type: \"paragraph\",\n      paragraph: {\n        rich_text: [],\n      },\n    });\n  }\n\n  return blocks;\n}\n\n/**\n * Extract Results Database ID from Notion page properties\n */\nfunction extractResultsDatabaseId(page: any): string | null {\n  const resultsDatabaseProperty = page.properties?.[\"Results Database ID\"];\n  if (!resultsDatabaseProperty || resultsDatabaseProperty.type !== \"rich_text\") {\n    return null;\n  }\n\n  if (resultsDatabaseProperty.rich_text?.length > 0) {\n    return resultsDatabaseProperty.rich_text[0].text.content;\n  }\n\n  return null;\n}\n\n/**\n * Extract Name property from Notion page\n */\nfunction extractNameProperty(page: any): string {\n  const nameProperty = page.properties?.Name;\n  if (!nameProperty) return \"\";\n\n  if (nameProperty.type === \"title\" && nameProperty.title?.length > 0) {\n    return nameProperty.title[0].text.content;\n  }\n\n  if (nameProperty.type === \"rich_text\" && nameProperty.rich_text?.length > 0) {\n    return nameProperty.rich_text[0].text.content;\n  }\n\n  return \"\";\n}\n\n/**\n * Extract Blob key property from Notion page\n */\nfunction extractBlobKeyProperty(page: any): string {\n  const blobKeyProperty = page.properties?.[\"Blob key\"];\n  if (!blobKeyProperty || blobKeyProperty.type !== \"rich_text\") return \"\";\n\n  if (blobKeyProperty.rich_text?.length > 0) {\n    return blobKeyProperty.rich_text[0].text.content;\n  }\n\n  return \"\";\n}\n\n/**\n * Create evidence summary text from per-transcript findings\n */\nfunction createEvidenceSummaryText(perTranscriptFindings: any[]): string {\n  if (perTranscriptFindings.length === 0) {\n    return \"\";\n  }\n\n  let summaryText = \"\";\n\n  for (const finding of perTranscriptFindings) {\n    const stanceEmoji =\n      finding.stance === \"supports\"\n        ? \"✅\"\n        : finding.stance === \"mixed\"\n        ? \"🟡\"\n        : finding.stance === \"contradicts\"\n        ? \"❌\"\n        : \"❓\";\n    summaryText += `- ${finding.key}: ${stanceEmoji} ${finding.stance} - ${finding.shortRationale}\\n`;\n  }\n\n  return summaryText;\n}\n",[],[235,241,266],{"_102":274,"_95":275,"_155":276,"_157":242,"_158":137,"_159":277,"_161":162,"_104":278,"_99":160,"_164":260,"_166":279,"_167":22,"_168":-5,"_169":280,"_98":281},"ee67af02-7dec-11f0-b478-0224a6c84d84","health.controller.ts","0198aa63-0ee2-725e-8c9f-92a15aa2aac7",32,["D",1755711701178],"import { getDatabaseById } from \"../services/notion.service.ts\";\n\nexport async function getConfiguredDatabases() {\n  const transcriptsDbId = Deno.env.get(\"FINDINGS_TRANSCRIPTS_DB_ID\");\n  const questionsDbId = Deno.env.get(\"FINDINGS_QUESTIONS_DB_ID\");\n  const reportsDbId = Deno.env.get(\"FINDINGS_REPORTS_RAW_DB_ID\");\n\n  const databases = {\n    transcripts: null,\n    questions: null,\n    reports: null,\n  };\n  const errors: string[] = [];\n\n  // Check transcripts database\n  if (transcriptsDbId) {\n    const result = await getDatabaseById(transcriptsDbId);\n    if (result.success) {\n      databases.transcripts = result.data;\n    } else {\n      errors.push(`Transcripts DB: ${result.error}`);\n    }\n  } else {\n    errors.push(\"FINDINGS_TRANSCRIPTS_DB_ID not configured\");\n  }\n\n  // Check questions database\n  if (questionsDbId) {\n    const result = await getDatabaseById(questionsDbId);\n    if (result.success) {\n      databases.questions = result.data;\n    } else {\n      errors.push(`Questions DB: ${result.error}`);\n    }\n  } else {\n    errors.push(\"FINDINGS_QUESTIONS_DB_ID not configured\");\n  }\n\n  // Check reports database\n  if (reportsDbId) {\n    const result = await getDatabaseById(reportsDbId);\n    if (result.success) {\n      databases.reports = result.data;\n    } else {\n      errors.push(`Reports DB: ${result.error}`);\n    }\n  } else {\n    errors.push(\"FINDINGS_REPORTS_RAW_DB_ID not configured\");\n  }\n\n  return {\n    success: errors.length === 0,\n    data: databases,\n    errors: errors.length > 0 ? errors : undefined,\n    timestamp: new Date().toISOString(),\n  };\n}\n\nexport async function getHealthStatus() {\n  const configuredDatabases = await getConfiguredDatabases();\n\n  const healthStatus = {\n    status: configuredDatabases.success ? \"ok\" : \"error\",\n    timestamp: new Date().toISOString(),\n    val: \"findings\",\n    databases: {\n      configured: configuredDatabases.success,\n      transcripts: configuredDatabases.data?.transcripts\n        ? \"connected\"\n        : \"error\",\n      questions: configuredDatabases.data?.questions ? \"connected\" : \"error\",\n      reports: configuredDatabases.data?.reports ? \"connected\" : \"error\",\n    },\n  };\n\n  if (configuredDatabases.errors) {\n    healthStatus.errors = configuredDatabases.errors;\n  }\n\n  return healthStatus;\n}\n",[],[235,241,275],{"_102":283,"_95":284,"_155":285,"_157":242,"_158":137,"_159":249,"_161":162,"_104":286,"_99":287,"_164":260,"_166":288,"_167":22,"_168":-5,"_169":289,"_98":290},"72e01910-8426-11f0-ad01-0224a6c84d84","tasks.controller.ts","c8fa40ba-79c6-11f0-95e0-0224a6c84d84",["D",1756396111921],5,"/**\n * Tasks Controller\n * Business logic for webhook-triggered task processing\n */\n\nimport { blob } from \"https://esm.town/v/std/blob\";\nimport {\n  getPageProperties,\n  updatePageStatus,\n  getPageBlocks,\n  createPageInDatabase,\n  copyBlocksToPage,\n  deleteAllDatabasePages,\n  createInlineDatabase,\n  createPageInDatabaseWithContent,\n  getDatabaseById,\n  getDatabasePages,\n  updatePageProperties,\n} from \"../services/notion.service.ts\";\nimport { extractTextFromPDFUrl } from \"../utils/pdfExtractor.ts\";\nimport { extractPageIdFromWebhook, extractFullNameFromPage } from \"../utils/notion-properties.utils.ts\";\nimport { redactPersonName } from \"../utils/text-redaction.utils.ts\";\nimport type {\n  DuplicatePageRequest,\n  DuplicatePageResponse,\n  ConvertPageRequest,\n  ConvertPageResponse,\n  DuplicatePageWithFilteredDatabaseRequest,\n  DuplicatePageWithFilteredDatabaseResponse,\n} from \"../types/findings.types.ts\";\n\nexport interface ProcessBlobRequest {\n  pageId: string;\n}\n\nexport interface ProcessBlobResponse {\n  success: boolean;\n  pageId: string;\n  blobKey?: string;\n  textLength?: number;\n  statusUpdated?: boolean;\n  timestamp: string;\n  error?: string;\n  details?: string;\n}\n\n/**\n * Process PDF extraction and blob storage for a Notion page\n */\nexport async function processBlobExtraction(\n  request: ProcessBlobRequest\n): Promise {\n  const { pageId } = request;\n  const timestamp = new Date().toISOString();\n\n  try {\n    console.log(`🔍 Processing page: ${pageId}`);\n\n    // Get page properties from Notion\n    const pageResult = await getPageProperties(pageId);\n    if (!pageResult.success) {\n      console.error(`❌ Failed to get page properties: ${pageResult.error}`);\n      return {\n        success: false,\n        pageId,\n        timestamp,\n        error: `Failed to get page properties: ${pageResult.error}`,\n      };\n    }\n\n    const page = pageResult.data;\n    console.log(`📄 Retrieved page properties for: ${page.id}`);\n\n    // Extract PDF URL from page properties\n    const pdfUrl = extractPdfUrl(page);\n    if (!pdfUrl) {\n      console.error(\"❌ No valid PDF found on page\");\n      return {\n        success: false,\n        pageId,\n        timestamp,\n        error: \"No PDF files found in PDF property\",\n      };\n    }\n\n    console.log(`📎 Found PDF URL: ${pdfUrl}`);\n\n    // Extract text from PDF\n    console.log(\"🔄 Starting PDF text extraction...\");\n    const extractedText = await extractTextFromPDFUrl(pdfUrl);\n\n    if (!extractedText || extractedText.trim().length === 0) {\n      console.error(\"❌ No text extracted from PDF\");\n      return {\n        success: false,\n        pageId,\n        timestamp,\n        error: \"No text could be extracted from PDF\",\n      };\n    }\n\n    // Extract person's name and redact from text\n    console.log(\"🔒 Redacting person's name from extracted text...\");\n    const fullName = extractFullNameFromPage(page);\n    const redactedText = fullName ? redactPersonName(extractedText, fullName) : extractedText;\n    \n    if (fullName) {\n      console.log(`✅ Redacted name \"${fullName}\" from text`);\n    } else {\n      console.log(\"⚠️ No name found to redact\");\n    }\n\n    // Save extracted text to blob storage\n    const blobKey = `findings--transcripts--${pageId}`;\n    console.log(`💾 Saving redacted text to blob with key: ${blobKey}`);\n\n    await blob.setJSON(blobKey, {\n      pageId: pageId,\n      extractedText: redactedText,\n      extractedAt: timestamp,\n      textLength: redactedText.length,\n      pdfUrl: pdfUrl,\n    });\n\n    console.log(`✅ Text saved to blob storage successfully`);\n\n    // Update page status to \"Done\" and save blob key\n    console.log(\"🔄 Updating page status to 'Done' and saving blob key...\");\n    const statusResult = await updatePageStatus(pageId, \"Done\", blobKey);\n    if (!statusResult.success) {\n      console.error(`⚠️ Failed to update page status: ${statusResult.error}`);\n      // Don't fail the entire operation if status update fails\n    } else {\n      console.log(\"✅ Page status updated to 'Done' and blob key saved\");\n    }\n\n    return {\n      success: true,\n      pageId,\n      blobKey,\n      textLength: redactedText.length,\n      statusUpdated: statusResult.success,\n      timestamp,\n    };\n  } catch (error) {\n    console.error(\"❌ Error processing blob extraction:\", error);\n    return {\n      success: false,\n      pageId,\n      timestamp,\n      error: \"Internal server error\",\n      details: error.message,\n    };\n  }\n}\n\n/**\n * Extract PDF URL from Notion page properties\n */\nfunction extractPdfUrl(page: any): string | null {\n  // Extract PDF property\n  const pdfProperty = page.properties?.PDF;\n  if (!pdfProperty || pdfProperty.type !== \"files\") {\n    console.error(\n      \"❌ No PDF property found or property is not of type 'files'\"\n    );\n    return null;\n  }\n\n  const files = pdfProperty.files;\n  if (!files || files.length === 0) {\n    console.error(\"❌ No files found in PDF property\");\n    return null;\n  }\n\n  // Get the first (and should be only) PDF file\n  const pdfFile = files[0];\n  if (!pdfFile) {\n    console.error(\"❌ PDF file is null or undefined\");\n    return null;\n  }\n\n  // Get the file URL (handle both external and Notion-hosted files)\n  if (pdfFile.type === \"external\") {\n    return pdfFile.external.url;\n  } else if (pdfFile.type === \"file\") {\n    return pdfFile.file.url;\n  } else {\n    console.error(\"❌ Unknown PDF file type:\", pdfFile.type);\n    return null;\n  }\n}\n\n/**\n * Duplicate a Notion page to the final reports database with content filtering\n */\nexport async function duplicatePageToFinalReports(\n  request: DuplicatePageRequest\n): Promise {\n  const { pageId } = request;\n  const timestamp = new Date().toISOString();\n\n  try {\n    console.log(`🔄 Starting page duplication for: ${pageId}`);\n\n    // Get environment variable for target database\n    const finalReportsDbId = Deno.env.get(\"FINDINGS_REPORTS_DRAFT_DB_ID\");\n    if (!finalReportsDbId) {\n      return {\n        success: false,\n        sourcePageId: pageId,\n        timestamp,\n        error:\n          \"FINDINGS_REPORTS_DRAFT_DB_ID environment variable not configured\",\n      };\n    }\n\n    // Step 1: Get source page properties\n    console.log(`📄 Getting source page properties...`);\n    const pageResult = await getPageProperties(pageId);\n    if (!pageResult.success) {\n      console.error(`❌ Failed to get page properties: ${pageResult.error}`);\n      return {\n        success: false,\n        sourcePageId: pageId,\n        timestamp,\n        error: `Failed to get source page properties: ${pageResult.error}`,\n      };\n    }\n\n    const sourcePage = pageResult.data;\n    console.log(`✅ Retrieved source page properties`);\n\n    // Step 2: Extract Name property for target page\n    const nameProperty = sourcePage.properties?.Name;\n    if (!nameProperty) {\n      console.error(`❌ No Name property found on source page`);\n      return {\n        success: false,\n        sourcePageId: pageId,\n        timestamp,\n        error: \"Source page has no Name property\",\n      };\n    }\n\n    // Step 3: Get source page content blocks\n    console.log(`📝 Getting source page content blocks...`);\n    const blocksResult = await getPageBlocks(pageId);\n    if (!blocksResult.success) {\n      console.error(`❌ Failed to get page blocks: ${blocksResult.error}`);\n      return {\n        success: false,\n        sourcePageId: pageId,\n        timestamp,\n        error: `Failed to get source page blocks: ${blocksResult.error}`,\n      };\n    }\n\n    const sourceBlocks = blocksResult.data.results;\n    console.log(`✅ Retrieved ${sourceBlocks.length} source blocks`);\n\n    // Step 4: Filter content blocks\n    console.log(`🔍 Filtering content blocks...`);\n    console.log(\n      `📋 Original blocks:`,\n      sourceBlocks.map((block) => ({ type: block.type, id: block.id }))\n    );\n\n    // Count H2 blocks before filtering\n    const h2CountBefore = sourceBlocks.filter(\n      (block) => block.type === \"heading_2\"\n    ).length;\n    console.log(`📊 Found ${h2CountBefore} H2 blocks in source`);\n\n    const filteredBlocks = filterPageBlocks(sourceBlocks);\n    const blocksFiltered = sourceBlocks.length - filteredBlocks.length;\n\n    // Count H2 blocks after filtering\n    const h2CountAfter = filteredBlocks.filter(\n      (block) => block.type === \"heading_2\"\n    ).length;\n    console.log(`📊 ${h2CountAfter} H2 blocks remaining after filtering`);\n\n    console.log(\n      `✅ Filtered ${blocksFiltered} blocks, ${filteredBlocks.length} blocks remaining`\n    );\n    console.log(\n      `📋 Remaining blocks:`,\n      filteredBlocks.map((block) => ({ type: block.type, id: block.id }))\n    );\n\n    // Step 5: Transform specific H3 blocks to bold paragraphs\n    console.log(`🔄 Transforming specific H3 blocks to bold paragraphs...`);\n    const transformedBlocks = transformPageBlocks(filteredBlocks);\n    const blocksTransformed =\n      filteredBlocks.length -\n      transformedBlocks.filter((block) => block.type === \"heading_3\").length +\n      transformedBlocks.filter(\n        (block) => block.type === \"paragraph\" && block._transformed\n      ).length;\n\n    console.log(\n      `✅ Transformed blocks, ${transformedBlocks.length} blocks total`\n    );\n    console.log(\n      `📋 Final blocks:`,\n      transformedBlocks.map((block) => ({\n        type: block.type,\n        id: block.id,\n        transformed: block._transformed || false,\n      }))\n    );\n\n    // Step 6: Create target page in final reports database\n    console.log(`📋 Creating target page in final reports database...`);\n\n    // Modify the Name property: replace \"(raw)\" with \"(draft)\"\n    let targetProperties;\n    if (\n      nameProperty.type === \"title\" &&\n      nameProperty.title &&\n      nameProperty.title.length > 0\n    ) {\n      // Get the current title text\n      const currentTitle = nameProperty.title[0].text.content;\n      const updatedTitle = currentTitle.replace(/\\(raw\\)/gi, \"(draft)\");\n\n      console.log(\n        `📝 Updating title from \"${currentTitle}\" to \"${updatedTitle}\"`\n      );\n\n      targetProperties = {\n        Name: {\n          title: [\n            {\n              type: \"text\",\n              text: {\n                content: updatedTitle,\n              },\n            },\n          ],\n        },\n      };\n    } else {\n      // Fallback: use original property if structure is unexpected\n      console.log(`⚠️ Unexpected Name property structure, using original`);\n      targetProperties = {\n        Name: nameProperty,\n      };\n    }\n\n    const createResult = await createPageInDatabase(\n      finalReportsDbId,\n      targetProperties\n    );\n    if (!createResult.success) {\n      console.error(`❌ Failed to create target page: ${createResult.error}`);\n      return {\n        success: false,\n        sourcePageId: pageId,\n        timestamp,\n        error: `Failed to create target page: ${createResult.error}`,\n      };\n    }\n\n    const targetPage = createResult.data;\n    const targetPageId = targetPage.id;\n    console.log(`✅ Created target page: ${targetPageId}`);\n\n    // Step 7: Copy filtered blocks to target page\n    if (transformedBlocks.length > 0) {\n      console.log(\n        `📋 Copying ${transformedBlocks.length} blocks to target page...`\n      );\n\n      // Clean blocks for copying (remove Notion-specific metadata and our transformation markers)\n      const cleanedBlocks = transformedBlocks.map((block) => {\n        const {\n          id,\n          created_time,\n          created_by,\n          last_edited_time,\n          last_edited_by,\n          archived,\n          has_children,\n          _transformed,\n          ...cleanBlock\n        } = block;\n        return cleanBlock;\n      });\n\n      console.log(\n        `📋 Cleaned blocks for copying:`,\n        cleanedBlocks.map((block) => ({ type: block.type }))\n      );\n\n      const copyResult = await copyBlocksToPage(targetPageId, cleanedBlocks);\n      if (!copyResult.success) {\n        console.error(`⚠️ Failed to copy blocks: ${copyResult.error}`);\n\n        // Check if it was a partial failure\n        if (copyResult.data?.successRate && copyResult.data.successRate >= 80) {\n          console.log(\n            `ℹ️ Partial success: ${copyResult.data.successRate}% of content was copied`\n          );\n          return {\n            success: true, // Mark as success since most content was copied\n            sourcePageId: pageId,\n            targetPageId: targetPageId,\n            blocksCopied: copyResult.data.totalBlocks || 0,\n            blocksFiltered: blocksFiltered,\n            timestamp,\n            warning:\n              copyResult.warning ||\n              `Partial copy: ${copyResult.data.successRate}% success rate`,\n          };\n        }\n\n        // Complete failure - create error content in target page\n        const errorBlocks = [\n          {\n            object: \"block\",\n            type: \"paragraph\",\n            paragraph: {\n              rich_text: [\n                {\n                  type: \"text\",\n                  text: {\n                    content: `Error copying content from source page: ${copyResult.error}`,\n                  },\n                },\n              ],\n            },\n          },\n        ];\n\n        await copyBlocksToPage(targetPageId, errorBlocks);\n\n        return {\n          success: false,\n          sourcePageId: pageId,\n          targetPageId: targetPageId,\n          blocksCopied: copyResult.data?.totalBlocks || 0,\n          blocksFiltered: blocksFiltered,\n          timestamp,\n          error: `Failed to copy blocks to target page: ${copyResult.error}`,\n        };\n      }\n      console.log(`✅ Successfully copied blocks to target page`);\n    } else {\n      console.log(`ℹ️ No blocks to copy after filtering and transformation`);\n    }\n\n    return {\n      success: true,\n      sourcePageId: pageId,\n      targetPageId: targetPageId,\n      blocksCopied: transformedBlocks.length,\n      blocksFiltered: blocksFiltered,\n      timestamp,\n    };\n  } catch (error) {\n    console.error(\"❌ Error duplicating page:\", error);\n    return {\n      success: false,\n      sourcePageId: pageId,\n      timestamp,\n      error: \"Internal server error\",\n      details: error.message,\n    };\n  }\n}\n\n/**\n * Filter page blocks to remove unwanted content\n */\nfunction filterPageBlocks(blocks: any[]): any[] {\n  return blocks.filter((block) => {\n    // Remove callout blocks\n    if (block.type === \"callout\") {\n      console.log(`🚫 Filtering out callout block`);\n      return false;\n    }\n\n    // Remove code blocks\n    if (block.type === \"code\") {\n      console.log(`🚫 Filtering out code block`);\n      return false;\n    }\n\n    // Remove divider blocks\n    if (block.type === \"divider\") {\n      console.log(`🚫 Filtering out divider block`);\n      return false;\n    }\n\n    // Remove blocks containing target phrases\n    const blockText = extractTextFromBlock(block);\n    if (blockText.toLowerCase().includes(\"evidence summary\")) {\n      console.log(`🚫 Filtering out block containing \"Evidence summary\"`);\n      return false;\n    }\n    if (blockText.toLowerCase().includes(\"methodology\")) {\n      console.log(`🚫 Filtering out block containing \"Methodology\"`);\n      return false;\n    }\n    if (blockText.toLowerCase().includes(\"executive summary\")) {\n      console.log(`🚫 Filtering out block containing \"Executive summary\"`);\n      return false;\n    }\n\n    return true;\n  });\n}\n\n/**\n * Transform specific H3 blocks to bold paragraphs\n */\nfunction transformPageBlocks(blocks: any[]): any[] {\n  return blocks.map((block) => {\n    // Only process heading_3 blocks\n    if (block.type !== \"heading_3\") {\n      return block;\n    }\n\n    // Extract text from the H3 block\n    const blockText = extractTextFromBlock(block);\n    const lowerText = blockText.toLowerCase().trim();\n\n    // Check if this H3 should be transformed\n    if (lowerText === \"key findings\" || lowerText === \"notable disagreements\") {\n      console.log(`🔄 Transforming H3 \"${blockText}\" to bold paragraph`);\n\n      // Create a new paragraph block with bold text\n      const transformedBlock = {\n        ...block,\n        type: \"paragraph\",\n        paragraph: {\n          rich_text: [\n            {\n              type: \"text\",\n              text: {\n                content: blockText,\n              },\n              annotations: {\n                bold: true,\n                italic: false,\n                strikethrough: false,\n                underline: false,\n                code: false,\n                color: \"default\",\n              },\n            },\n          ],\n        },\n        _transformed: true, // Mark for logging purposes\n      };\n\n      // Remove the original heading_3 property\n      delete transformedBlock.heading_3;\n\n      return transformedBlock;\n    }\n\n    // Return unchanged if not a target heading\n    return block;\n  });\n}\n\n/**\n * Extract text content from various Notion block types\n */\nfunction extractTextFromBlock(block: any): string {\n  if (!block || !block.type) return \"\";\n\n  const blockType = block[block.type];\n  if (!blockType) return \"\";\n\n  // Handle blocks with rich_text property\n  if (blockType.rich_text && Array.isArray(blockType.rich_text)) {\n    return blockType.rich_text\n      .map((text: any) => text.plain_text || text.text?.content || \"\")\n      .join(\"\");\n  }\n\n  // Handle blocks with text property (some block types use this)\n  if (blockType.text && Array.isArray(blockType.text)) {\n    return blockType.text\n      .map((text: any) => text.plain_text || text.text?.content || \"\")\n      .join(\"\");\n  }\n\n  // Handle title property (for some block types)\n  if (blockType.title && Array.isArray(blockType.title)) {\n    return blockType.title\n      .map((text: any) => text.plain_text || text.text?.content || \"\")\n      .join(\"\");\n  }\n\n  return \"\";\n}\n\n/**\n * Convert a Notion page to database pages based on H2 sections\n */\nexport async function convertPageToDatabase(\n  request: ConvertPageRequest\n): Promise {\n  const { pageId } = request;\n  const timestamp = new Date().toISOString();\n  const warnings: string[] = [];\n\n  try {\n    console.log(`🔄 Starting page conversion for: ${pageId}`);\n\n    // Get environment variable for target database\n    const targetDatabaseId = Deno.env.get(\"FINDINGS_REPORT_FINAL_DB_ID\");\n    if (!targetDatabaseId) {\n      return {\n        success: false,\n        sourcePageId: pageId,\n        targetDatabaseId: \"\",\n        h2SectionsFound: 0,\n        pagesCreated: 0,\n        pagesWithContent: 0,\n        timestamp,\n        error:\n          \"FINDINGS_REPORT_FINAL_DB_ID environment variable not configured\",\n      };\n    }\n\n    // Step 1: Get source page content blocks\n    console.log(`📝 Getting source page content blocks...`);\n    const blocksResult = await getPageBlocks(pageId);\n    if (!blocksResult.success) {\n      console.error(`❌ Failed to get page blocks: ${blocksResult.error}`);\n      return {\n        success: false,\n        sourcePageId: pageId,\n        targetDatabaseId,\n        h2SectionsFound: 0,\n        pagesCreated: 0,\n        pagesWithContent: 0,\n        timestamp,\n        error: `Failed to get source page blocks: ${blocksResult.error}`,\n      };\n    }\n\n    const sourceBlocks = blocksResult.data.results;\n    console.log(`✅ Retrieved ${sourceBlocks.length} source blocks`);\n\n    // Debug: Log all block types to see what we're working with\n    console.log(\n      `🔍 Block types found:`,\n      sourceBlocks.map((block) => ({ type: block.type, id: block.id }))\n    );\n\n    // Step 2: Parse blocks to identify H2 sections\n    console.log(`🔍 Parsing H2 sections...`);\n    const h2Sections = parseH2Sections(sourceBlocks);\n    console.log(`✅ Found ${h2Sections.length} H2 sections`);\n\n    if (h2Sections.length === 0) {\n      return {\n        success: false,\n        sourcePageId: pageId,\n        targetDatabaseId,\n        h2SectionsFound: 0,\n        pagesCreated: 0,\n        pagesWithContent: 0,\n        timestamp,\n        error: \"No H2 sections found in source page\",\n      };\n    }\n\n    // Step 3: Delete all existing pages from target database\n    console.log(`🗑️ Deleting all pages from target database...`);\n    const deleteResult = await deleteAllDatabasePages(targetDatabaseId);\n    if (!deleteResult.success) {\n      console.error(\n        `❌ Failed to delete database pages: ${deleteResult.error}`\n      );\n      return {\n        success: false,\n        sourcePageId: pageId,\n        targetDatabaseId,\n        h2SectionsFound: h2Sections.length,\n        pagesCreated: 0,\n        pagesWithContent: 0,\n        timestamp,\n        error: `Failed to delete existing database pages: ${deleteResult.error}`,\n      };\n    }\n\n    if (deleteResult.warning) {\n      warnings.push(deleteResult.warning);\n    }\n\n    console.log(\n      `✅ Deleted ${deleteResult.data.deletedPages} pages from database`\n    );\n\n    // Step 4: Create new pages for each H2 section\n    let pagesCreated = 0;\n    let pagesWithContent = 0;\n\n    for (let i = 0; i < h2Sections.length; i++) {\n      const section = h2Sections[i];\n      console.log(\n        `📋 Processing H2 section ${i + 1}/${h2Sections.length}: \"${\n          section.title\n        }\"`\n      );\n\n      try {\n        // Create page with H2 title as Name property\n        const pageProperties = {\n          Name: {\n            title: [\n              {\n                type: \"text\",\n                text: {\n                  content: section.title,\n                },\n              },\n            ],\n          },\n        };\n\n        const createResult = await createPageInDatabase(\n          targetDatabaseId,\n          pageProperties\n        );\n        if (!createResult.success) {\n          console.error(\n            `❌ Failed to create page for \"${section.title}\": ${createResult.error}`\n          );\n          warnings.push(\n            `Failed to create page for \"${section.title}\": ${createResult.error}`\n          );\n          continue;\n        }\n\n        const newPageId = createResult.data.id;\n        pagesCreated++;\n        console.log(\n          `✅ Created page: ${newPageId} with title \"${section.title}\"`\n        );\n\n        // Copy content blocks to the new page (if any)\n        if (section.contentBlocks.length > 0) {\n          console.log(\n            `📋 Copying ${section.contentBlocks.length} content blocks...`\n          );\n\n          // Clean blocks for copying (remove Notion-specific metadata)\n          const cleanedBlocks = section.contentBlocks.map((block) => {\n            const {\n              id,\n              created_time,\n              created_by,\n              last_edited_time,\n              last_edited_by,\n              archived,\n              has_children,\n              ...cleanBlock\n            } = block;\n            return cleanBlock;\n          });\n\n          const copyResult = await copyBlocksToPage(newPageId, cleanedBlocks);\n          if (!copyResult.success) {\n            console.error(\n              `⚠️ Failed to copy content to \"${section.title}\": ${copyResult.error}`\n            );\n            warnings.push(\n              `Failed to copy content to \"${section.title}\": ${copyResult.error}`\n            );\n          } else {\n            pagesWithContent++;\n            console.log(`✅ Successfully copied content to \"${section.title}\"`);\n          }\n        } else {\n          console.log(`ℹ️ No content blocks to copy for \"${section.title}\"`);\n        }\n\n        // Small delay to avoid rate limiting\n        await new Promise((resolve) => setTimeout(resolve, 200));\n      } catch (sectionError) {\n        console.error(\n          `❌ Error processing section \"${section.title}\":`,\n          sectionError\n        );\n        warnings.push(\n          `Error processing section \"${section.title}\": ${sectionError.message}`\n        );\n      }\n    }\n\n    // Determine overall success\n    const successRate = pagesCreated / h2Sections.length;\n    const isSuccess = successRate >= 0.8; // Consider success if 80%+ of sections were processed\n\n    console.log(\n      `📊 Conversion complete: ${pagesCreated}/${h2Sections.length} pages created, ${pagesWithContent} with content`\n    );\n\n    return {\n      success: isSuccess,\n      sourcePageId: pageId,\n      targetDatabaseId,\n      h2SectionsFound: h2Sections.length,\n      pagesCreated,\n      pagesWithContent,\n      timestamp,\n      warnings: warnings.length > 0 ? warnings : undefined,\n    };\n  } catch (error) {\n    console.error(\"❌ Error converting page to database:\", error);\n    return {\n      success: false,\n      sourcePageId: pageId,\n      targetDatabaseId: \"\",\n      h2SectionsFound: 0,\n      pagesCreated: 0,\n      pagesWithContent: 0,\n      timestamp,\n      error: \"Internal server error\",\n      details: error.message,\n    };\n  }\n}\n\n/**\n * Parse blocks to identify H2 sections and their content\n */\nfunction parseH2Sections(\n  blocks: any[]\n): Array<{ title: string; contentBlocks: any[] }> {\n  const sections: Array<{ title: string; contentBlocks: any[] }> = [];\n  let currentSection: { title: string; contentBlocks: any[] } | null = null;\n\n  console.log(\n    `🔍 Starting to parse ${blocks.length} blocks for H2 sections...`\n  );\n\n  for (let i = 0; i < blocks.length; i++) {\n    const block = blocks[i];\n    console.log(\n      `📋 Processing block ${i + 1}: type=\"${block.type}\", id=\"${block.id}\"`\n    );\n\n    if (block.type === \"heading_2\") {\n      // Save previous section if it exists\n      if (currentSection) {\n        console.log(\n          `💾 Saving previous section \"${currentSection.title}\" with ${currentSection.contentBlocks.length} content blocks`\n        );\n        sections.push(currentSection);\n      }\n\n      // Start new section with H2 title (exclude the H2 block itself from content)\n      const title = extractTextFromBlock(block);\n      currentSection = {\n        title: title || \"Untitled Section\",\n        contentBlocks: [],\n      };\n\n      console.log(`📝 Found H2 section: \"${currentSection.title}\"`);\n    } else if (currentSection) {\n      // Add content block to current section (excluding the H2 itself)\n      console.log(\n        `📄 Adding content block to section \"${currentSection.title}\": type=\"${block.type}\"`\n      );\n      currentSection.contentBlocks.push(block);\n    } else {\n      console.log(`⏭️ Skipping block before first H2: type=\"${block.type}\"`);\n    }\n    // Ignore blocks before the first H2 (as per requirements)\n  }\n\n  // Don't forget the last section\n  if (currentSection) {\n    console.log(\n      `💾 Saving final section \"${currentSection.title}\" with ${currentSection.contentBlocks.length} content blocks`\n    );\n    sections.push(currentSection);\n  }\n\n  console.log(`✅ Parsing complete: found ${sections.length} H2 sections`);\n  return sections;\n}\n\n/**\n * Duplicate a Notion page with its child database, filtering specific properties\n */\nexport async function duplicatePageWithFilteredDatabase(\n  request: DuplicatePageWithFilteredDatabaseRequest\n): Promise {\n  const { pageId } = request;\n  const timestamp = new Date().toISOString();\n  const warnings: string[] = [];\n\n  try {\n    console.log(`🔄 Starting page duplication with filtered database for: ${pageId}`);\n\n    // Get environment variable for target database\n    const targetDatabaseId = Deno.env.get(\"FINDINGS_REPORTS_DRAFT_DB_ID\");\n    if (!targetDatabaseId) {\n      return {\n        success: false,\n        sourcePageId: pageId,\n        timestamp,\n        error: \"FINDINGS_REPORTS_DRAFT_DB_ID environment variable not configured\",\n      };\n    }\n\n    // Step 1: Get source page properties\n    console.log(`📄 Getting source page properties...`);\n    const pageResult = await getPageProperties(pageId);\n    if (!pageResult.success) {\n      console.error(`❌ Failed to get page properties: ${pageResult.error}`);\n      return {\n        success: false,\n        sourcePageId: pageId,\n        timestamp,\n        error: `Failed to get source page properties: ${pageResult.error}`,\n      };\n    }\n\n    const sourcePage = pageResult.data;\n    console.log(`✅ Retrieved source page properties`);\n\n    // Step 2: Extract child database ID from \"Results Database ID\" property\n    const childDatabaseId = extractResultsDatabaseId(sourcePage);\n    if (!childDatabaseId) {\n      console.log(`⚠️ No child database found in \"Results Database ID\" property`);\n      warnings.push(\"No child database found to duplicate\");\n    }\n\n    // Step 3: Extract Name property for target page\n    const nameProperty = sourcePage.properties?.Name;\n    if (!nameProperty) {\n      console.error(`❌ No Name property found on source page`);\n      return {\n        success: false,\n        sourcePageId: pageId,\n        timestamp,\n        error: \"Source page has no Name property\",\n      };\n    }\n\n    // Step 4: Get source page content blocks\n    console.log(`📝 Getting source page content blocks...`);\n    const blocksResult = await getPageBlocks(pageId);\n    if (!blocksResult.success) {\n      console.error(`❌ Failed to get page blocks: ${blocksResult.error}`);\n      return {\n        success: false,\n        sourcePageId: pageId,\n        timestamp,\n        error: `Failed to get source page blocks: ${blocksResult.error}`,\n      };\n    }\n\n    const sourceBlocks = blocksResult.data.results;\n    console.log(`✅ Retrieved ${sourceBlocks.length} source blocks`);\n\n    // Step 5: Create target page in draft reports database\n    console.log(`📋 Creating target page in draft reports database...`);\n\n    // Modify the Name property: replace \"(raw)\" with \"(draft)\" if present\n    let targetProperties;\n    if (\n      nameProperty.type === \"title\" &&\n      nameProperty.title &&\n      nameProperty.title.length > 0\n    ) {\n      const currentTitle = nameProperty.title[0].text.content;\n      const updatedTitle = currentTitle.includes(\"(raw)\") \n        ? currentTitle.replace(/\\\\(raw\\\\)/gi, \"(draft)\")\n        : `${currentTitle} (draft)`;\n\n      console.log(\n        `📝 Updating title from \"${currentTitle}\" to \"${updatedTitle}\"`\n      );\n\n      targetProperties = {\n        Name: {\n          title: [\n            {\n              type: \"text\",\n              text: {\n                content: updatedTitle,\n              },\n            },\n          ],\n        },\n      };\n    } else {\n      console.log(`⚠️ Unexpected Name property structure, using original`);\n      targetProperties = {\n        Name: nameProperty,\n      };\n    }\n\n    const createResult = await createPageInDatabase(\n      targetDatabaseId,\n      targetProperties\n    );\n    if (!createResult.success) {\n      console.error(`❌ Failed to create target page: ${createResult.error}`);\n      return {\n        success: false,\n        sourcePageId: pageId,\n        timestamp,\n        error: `Failed to create target page: ${createResult.error}`,\n      };\n    }\n\n    const targetPage = createResult.data;\n    const targetPageId = targetPage.id;\n    console.log(`✅ Created target page: ${targetPageId}`);\n\n    let targetChildDatabaseId: string | undefined;\n    let childPagesProcessed = 0;\n    let childPagesSucceeded = 0;\n\n    // Step 6: Handle child database duplication if it exists\n    if (childDatabaseId) {\n      console.log(`📊 Processing child database: ${childDatabaseId}`);\n\n      try {\n        // Get child database schema\n        console.log(`🔍 Getting child database schema...`);\n        const childDbResult = await getDatabaseById(childDatabaseId);\n        if (!childDbResult.success) {\n          console.error(`❌ Failed to get child database: ${childDbResult.error}`);\n          warnings.push(`Failed to get child database: ${childDbResult.error}`);\n        } else {\n          // Create filtered schema (exclude specific properties)\n          const originalSchema = childDbResult.data.properties;\n          const filteredSchema = createFilteredDatabaseSchema(originalSchema);\n          \n          console.log(`📊 Creating filtered child database in target page...`);\n          const createChildDbResult = await createInlineDatabase(\n            targetPageId,\n            \"Analysis Results\",\n            filteredSchema\n          );\n\n          if (!createChildDbResult.success) {\n            console.error(`❌ Failed to create child database: ${createChildDbResult.error}`);\n            warnings.push(`Failed to create child database: ${createChildDbResult.error}`);\n          } else {\n            targetChildDatabaseId = createChildDbResult.data.databaseId;\n            console.log(`✅ Created filtered child database: ${targetChildDatabaseId}`);\n\n            // Get all pages from source child database\n            console.log(`📋 Getting pages from source child database...`);\n            const childPagesResult = await getDatabasePages(childDatabaseId);\n            if (!childPagesResult.success) {\n              console.error(`❌ Failed to get child database pages: ${childPagesResult.error}`);\n              warnings.push(`Failed to get child database pages: ${childPagesResult.error}`);\n            } else {\n              const childPages = childPagesResult.data.results;\n              childPagesProcessed = childPages.length;\n              console.log(`📄 Found ${childPages.length} pages in child database`);\n\n              // Duplicate each child page with filtered properties and content\n              for (const childPage of childPages) {\n                try {\n                  console.log(`📄 Processing child page: ${childPage.id}`);\n                  \n                  // Get child page content blocks\n                  const childBlocksResult = await getPageBlocks(childPage.id);\n                  let childContentBlocks = [];\n                  \n                  if (childBlocksResult.success) {\n                    // Clean blocks for copying (remove Notion metadata)\n                    childContentBlocks = childBlocksResult.data.results.map((block) => {\n                      const {\n                        id,\n                        created_time,\n                        created_by,\n                        last_edited_time,\n                        last_edited_by,\n                        archived,\n                        has_children,\n                        ...cleanBlock\n                      } = block;\n                      return cleanBlock;\n                    });\n                    console.log(`📋 Retrieved ${childContentBlocks.length} content blocks for child page`);\n                  } else {\n                    console.error(`⚠️ Failed to get content blocks for child page ${childPage.id}: ${childBlocksResult.error}`);\n                    warnings.push(`Failed to get content blocks for child page: ${childBlocksResult.error}`);\n                  }\n\n                  const filteredProperties = filterPageProperties(childPage.properties, filteredSchema);\n                  const createChildPageResult = await createPageInDatabaseWithContent(\n                    targetChildDatabaseId,\n                    filteredProperties,\n                    childContentBlocks // Include the content blocks\n                  );\n\n                  if (createChildPageResult.success) {\n                    childPagesSucceeded++;\n                    console.log(`✅ Duplicated child page: ${childPage.id}`);\n                  } else {\n                    console.error(`❌ Failed to duplicate child page ${childPage.id}: ${createChildPageResult.error}`);\n                    warnings.push(`Failed to duplicate child page: ${createChildPageResult.error}`);\n                  }\n                } catch (childPageError) {\n                  console.error(`❌ Error duplicating child page ${childPage.id}:`, childPageError);\n                  warnings.push(`Error duplicating child page: ${childPageError.message}`);\n                }\n              }\n\n              console.log(`📊 Child database duplication complete: ${childPagesSucceeded}/${childPagesProcessed} pages succeeded`);\n            }\n          }\n        }\n      } catch (childDbError) {\n        console.error(`❌ Error processing child database:`, childDbError);\n        warnings.push(`Error processing child database: ${childDbError.message}`);\n      }\n    }\n\n    // Step 7: Copy source page content blocks (excluding child database blocks)\n    if (sourceBlocks.length > 0) {\n      console.log(`📋 Copying ${sourceBlocks.length} content blocks to target page...`);\n\n      // Filter out child database blocks if they exist\n      const filteredBlocks = sourceBlocks.filter(block => {\n        if (block.type === \"child_database\") {\n          console.log(`🚫 Filtering out child_database block`);\n          return false;\n        }\n        return true;\n      });\n\n      if (filteredBlocks.length > 0) {\n        // Clean blocks for copying\n        const cleanedBlocks = filteredBlocks.map((block) => {\n          const {\n            id,\n            created_time,\n            created_by,\n            last_edited_time,\n            last_edited_by,\n            archived,\n            has_children,\n            ...cleanBlock\n          } = block;\n          return cleanBlock;\n        });\n\n        const copyResult = await copyBlocksToPage(targetPageId, cleanedBlocks);\n        if (!copyResult.success) {\n          console.error(`⚠️ Failed to copy blocks: ${copyResult.error}`);\n          warnings.push(`Failed to copy content blocks: ${copyResult.error}`);\n        } else {\n          console.log(`✅ Successfully copied ${filteredBlocks.length} content blocks`);\n        }\n      } else {\n        console.log(`ℹ️ No content blocks to copy after filtering`);\n      }\n    }\n\n    const successRate = childDatabaseId ? (childPagesSucceeded / Math.max(childPagesProcessed, 1)) : 1;\n    const isSuccess = successRate >= 0.8; // Consider success if 80%+ of child pages were processed\n\n    return {\n      success: isSuccess,\n      sourcePageId: pageId,\n      targetPageId: targetPageId,\n      childDatabaseId: childDatabaseId,\n      targetChildDatabaseId: targetChildDatabaseId,\n      childPagesProcessed: childPagesProcessed,\n      childPagesSucceeded: childPagesSucceeded,\n      timestamp,\n      warnings: warnings.length > 0 ? warnings : undefined,\n    };\n  } catch (error) {\n    console.error(\"❌ Error duplicating page with filtered database:\", error);\n    return {\n      success: false,\n      sourcePageId: pageId,\n      timestamp,\n      error: \"Internal server error\",\n      warnings: [`Internal error: ${error.message}`],\n    };\n  }\n}\n\n/**\n * Extract Results Database ID from Notion page properties\n */\nfunction extractResultsDatabaseId(page: any): string | null {\n  const resultsDatabaseProperty = page.properties?.[\"Results Database ID\"];\n  if (!resultsDatabaseProperty || resultsDatabaseProperty.type !== \"rich_text\") {\n    return null;\n  }\n\n  if (resultsDatabaseProperty.rich_text?.length > 0) {\n    return resultsDatabaseProperty.rich_text[0].text.content;\n  }\n\n  return null;\n}\n\n/**\n * Create filtered database schema excluding specific properties\n */\nfunction createFilteredDatabaseSchema(originalSchema: any): any {\n  const excludedProperties = [\n    \"Confidence Score\",\n    \"Consensus Level\", \n    \"Evidence Summary\",\n    \"Methodology\"\n  ];\n\n  const filteredSchema: any = {};\n  \n  for (const [propertyName, propertyConfig] of Object.entries(originalSchema)) {\n    if (!excludedProperties.includes(propertyName)) {\n      filteredSchema[propertyName] = propertyConfig;\n    } else {\n      console.log(`🚫 Excluding property: ${propertyName}`);\n    }\n  }\n\n  console.log(`📊 Filtered schema: ${Object.keys(filteredSchema).length} properties (excluded ${excludedProperties.length})`);\n  return filteredSchema;\n}\n\n/**\n * Filter page properties excluding specific fields and handle select options\n */\nfunction filterPageProperties(originalProperties: any, targetSchema: any): any {\n  const excludedProperties = [\n    \"Confidence Score\",\n    \"Consensus Level\", \n    \"Evidence Summary\",\n    \"Methodology\"\n  ];\n\n  const filteredProperties: any = {};\n  \n  for (const [propertyName, propertyValue] of Object.entries(originalProperties)) {\n    if (!excludedProperties.includes(propertyName) && targetSchema[propertyName]) {\n      // Handle select properties specially to avoid ID mismatches\n      if (propertyValue && typeof propertyValue === 'object' && 'select' in propertyValue) {\n        const selectValue = propertyValue.select;\n        if (selectValue && selectValue.name) {\n          // Use name instead of ID to avoid option ID mismatches\n          filteredProperties[propertyName] = {\n            select: {\n              name: selectValue.name\n            }\n          };\n        } else {\n          filteredProperties[propertyName] = {\n            select: null\n          };\n        }\n      } else if (propertyValue && typeof propertyValue === 'object' && 'multi_select' in propertyValue) {\n        const multiSelectValue = propertyValue.multi_select;\n        if (multiSelectValue && Array.isArray(multiSelectValue)) {\n          // Use names instead of IDs for multi-select\n          filteredProperties[propertyName] = {\n            multi_select: multiSelectValue.map((option: any) => ({\n              name: option.name\n            }))\n          };\n        } else {\n          filteredProperties[propertyName] = {\n            multi_select: []\n          };\n        }\n      } else {\n        // For other property types, copy as-is\n        filteredProperties[propertyName] = propertyValue;\n      }\n    }\n  }\n\n  return filteredProperties;\n}\n",[],[235,241,284],[235,241],{"_102":293,"_95":294,"_155":295,"_157":236,"_158":137,"_159":160,"_161":162,"_104":296,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_169":297,"_98":316},"d0f9bc20-7951-11f0-8810-0224a6c84d84","crons","0198aa63-0ee2-725e-8c9f-9f1fa6f65456",["D",1755205275298],[298,307],{"_102":299,"_95":247,"_155":300,"_157":295,"_158":137,"_159":301,"_161":162,"_104":302,"_99":303,"_164":177,"_166":304,"_167":22,"_168":-5,"_169":305,"_98":306},"24e30a3c-7a18-11f0-8841-0224a6c84d84","0198aa63-0ee2-725e-8c9f-9afb0718c9af",24,["D",1755290456490],2,"# Crons\n\nScheduled task handlers.\n\n## Available Crons\n\n- `findings-analysis.ts` - **Batched Processing**: Analyzes transcripts using enhanced two-stage consensus analysis and generates rich findings reports with:\n  - **Batch Size**: Processes 5 questions per cron run to avoid timeouts\n  - **State Management**: Uses question status tracking (\"Not started\" → \"Working...\" → \"Done\"/\"Error\")\n  - **Resume Logic**: Automatically continues from where previous runs left off\n  - **Per-transcript individual analysis**\n  - **Cross-transcript consensus synthesis**  \n  - **Confidence scores and consensus levels**\n  - **Disagreement tracking and evidence summaries**\n  - **Rich Notion formatting with callouts, tables, and structured sections**\n\n### Batched Processing Workflow\n\n1. **Initialization**: When a new report is found, all questions are reset to \"Not started\"\n2. **Batch Selection**: Each cron run picks up to 5 questions with \"Not started\" status\n3. **Processing**: Selected questions are marked \"Working...\" then processed individually\n4. **Completion**: Successfully processed questions are marked \"Done\", failed ones marked \"Error\"\n5. **Resume**: Subsequent cron runs continue with remaining \"Not started\" questions\n6. **Finalization**: When all questions are complete, the report is marked \"Done\"\n\n### Timing\n- **Per Question**: ~30 seconds average processing time\n- **Per Batch**: ~3-4 minutes total (5 questions + overhead)\n- **Safety Margin**: Well under timeout limits with room for variability",[],[235,294,247],{"_102":308,"_95":309,"_155":310,"_157":295,"_158":137,"_159":301,"_161":162,"_104":311,"_99":160,"_164":312,"_166":313,"_167":22,"_168":-5,"_169":314,"_98":315},"24e30d8e-7a18-11f0-8841-0224a6c84d84","findings-analysis.ts","82c45a60-7a05-11f0-9236-0224a6c84d84",["D",1755290456490],"interval","/**\n * Findings Analysis Cron Job\n * Scheduled task to analyze transcripts and generate findings reports in batches\n */\n\nimport { processFindings } from \"../controllers/findings.controller.ts\";\n\nexport default async function() {\n  console.log(\"🚀 Starting findings analysis cron job (batched processing)...\");\n  \n  const result = await processFindings({});\n  \n  if (result.success) {\n    if (result.remainingQuestions !== undefined && result.remainingQuestions > 0) {\n      // Batch completed, more work remaining\n      console.log(\"✅ Batch processing completed successfully\");\n      console.log(`📊 Report page: ${result.reportPageId}`);\n      console.log(`❓ Questions in this batch: ${result.questionsProcessed}`);\n      console.log(`✅ Questions succeeded: ${result.questionsSucceeded}`);\n      console.log(`⏳ Questions remaining: ${result.remainingQuestions}`);\n      console.log(\"🔄 Next cron run will process the remaining questions\");\n      \n      if (result.questionsSucceeded !== result.questionsProcessed) {\n        console.log(`⚠️ Some questions in this batch failed: ${result.questionsProcessed - result.questionsSucceeded} errors`);\n      }\n    } else {\n      // All processing complete\n      console.log(\"🎉 All findings analysis completed successfully!\");\n      console.log(`📊 Report page: ${result.reportPageId}`);\n      console.log(`❓ Total questions processed: ${result.questionsProcessed}`);\n      console.log(`✅ Questions succeeded: ${result.questionsSucceeded}`);\n      \n      if (result.questionsSucceeded !== result.questionsProcessed) {\n        console.log(`⚠️ Some questions failed: ${result.questionsProcessed - result.questionsSucceeded} errors`);\n      }\n    }\n  } else {\n    console.error(\"❌ Findings analysis failed:\", result.error);\n    if (result.details) {\n      console.error(\"Details:\", result.details);\n    }\n  }\n  \n  return result;\n}",[],[235,294,309],[235,294],{"_102":318,"_95":319,"_155":320,"_157":236,"_158":137,"_159":218,"_161":162,"_104":321,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_169":322,"_98":340},"fdca9eb6-841b-11f0-bd95-0224a6c84d84","emails","0198f119-3f9f-7273-a5ea-5547629990ee",["D",1756391620516],[323,330],{"_102":324,"_95":247,"_155":325,"_157":320,"_158":137,"_159":136,"_161":162,"_104":326,"_99":303,"_164":177,"_166":327,"_167":22,"_168":-5,"_169":328,"_98":329},"e2cc6dc8-845d-11f0-a352-0224a6c84d84","fdca71a2-841b-11f0-bd95-0224a6c84d84",["D",1756419922020],"# Emails\n\nEmail trigger handlers for processing incoming emails.\n\n## Files\n\n- `pdf-attachments.ts` - Processes emails with PDF attachments and uploads them to Notion database pages\n\n## Architecture\n\nEmail handlers follow the same separation of concerns as other parts of the application:\n\n- **Email Handlers**: Extract data from email objects and call controllers\n- **Controllers**: Business logic for processing email data\n- **Services**: Pure API calls to external systems (Notion, etc.)\n\n## PDF Attachments Handler\n\nThe `pdf-attachments.ts` handler:\n\n1. **Deduplication Check**: Uses precise time difference comparison to prevent duplicate processing\n2. Receives emails with PDF attachments\n3. Creates child pages in the Notion database specified by `FINDINGS_TRANSCRIPTS_DB_ID`\n4. Uploads each PDF to its own page using Notion's 3-step file upload API\n5. Continues processing even if individual uploads fail\n6. Returns detailed results for monitoring\n\n### Duplicate Email Protection\n\nThe handler implements idempotency using precise time difference comparison:\n\n- **Hash Generation**: `from + subject` (no time component)\n- **Storage**: Blob storage with key `findings--transcripts--email--{hash}`\n- **Window**: Same sender + subject within 3 minutes = duplicate\n- **Precision**: Exact time difference calculation (not epoch blocks)\n- **Behavior**: Duplicates are logged with time difference and skipped\n- **Updates**: Legitimate emails (>3 minutes apart) update the stored timestamp\n\nThis prevents duplicate Notion records when email providers retry delivery while allowing legitimate re-sends after 3 minutes.\n\n### Page Properties\n\nEach PDF attachment creates a page with:\n\n- **Name**: Extracted interviewee name from filename (content in parentheses)\n- **Filename**: Complete original filename from email attachment\n- **PDF**: Uploaded file using Notion's file upload API\n- **Status**: Upload progress tracking\n\n### Status Progression\n\n- **\"Uploading...\"**: Initial status when page is created\n- **\"Uploaded\"**: File successfully uploaded and attached to page\n- **\"Upload error\"**: File upload failed (page still exists for retry)\n\n### Filename Processing\n\nThe handler extracts interviewee names using pattern matching:\n\n- **Pattern**: `(Name)` - Content inside first parentheses becomes the page title\n- **Example**: `\"Interview (John Doe) - 2025_01_01.pdf\"` → Name: `\"John Doe\"`\n- **Fallback**: If no parentheses found, uses filename without extension\n\n### Environment Variables\n\n- `FINDINGS_TRANSCRIPTS_DB_ID` - Notion database ID where pages will be created\n- `NOTION_API_KEY` - Notion API key for authentication\n\n### Process Flow\n\n1. **Email Reception**: Handler receives email with attachments\n2. **PDF Filtering**: Only processes PDF attachments (by extension or MIME type)\n3. **Page Creation**: Creates a new page in the target database for each PDF\n4. **File Upload**: Uses Notion's 3-step upload process:\n   - Create upload object\n   - Upload file contents\n   - Attach file to page\n5. **Status Updates**: Updates page status based on upload success/failure\n6. **Error Handling**: Continues processing even if individual steps fail\n7. **Response**: Returns detailed processing results",[],[235,319,247],{"_102":331,"_95":332,"_155":333,"_157":320,"_158":137,"_159":334,"_161":162,"_104":335,"_99":160,"_164":336,"_166":337,"_167":22,"_168":96,"_169":338,"_98":339},"ca09c2e0-845d-11f0-a3c1-0224a6c84d84","pdf-attachments.ts","fdca6c7a-841b-11f0-bd95-0224a6c84d84",45,["D",1756419880478],"email","/**\n * Email Handler: PDF Attachments to Notion\n * Orchestrates email processing workflow with proper separation of concerns\n * \n * Responsibilities:\n * - Validate incoming emails\n * - Extract PDF attachments\n * - Coordinate with email service for user communication\n * - Trigger business logic processing\n * - Return quick response to prevent email service timeouts\n */\n\nimport { blob } from \"https://esm.town/v/std/blob\";\nimport { processEmailPdfAttachments } from \"../controllers/emails.controller.ts\";\nimport { sendProcessingAcknowledgment } from \"../services/email.service.ts\";\nimport type { Email } from \"../types/email.types.ts\";\n\nexport default async function(email: Email) {\n  const processingId = `${Date.now()}-${Math.random().toString(36).substr(2, 9)}`;\n  \n  try {\n    console.log(`📧 [${processingId}] Received email: ${email.subject || 'No Subject'}`);\n    console.log(`📧 [${processingId}] From: ${email.from}`);\n    console.log(`📧 [${processingId}] Attachments: ${email.attachments?.length || 0}`);\n\n    // Email deduplication check using 3-minute time difference\n    const currentEmailTime = email.date ? new Date(email.date).getTime() : Date.now();\n    const hashInput = `${email.from}|${email.subject || 'No Subject'}`;\n    const emailHash = await crypto.subtle.digest('SHA-256', new TextEncoder().encode(hashInput))\n      .then(buffer => Array.from(new Uint8Array(buffer))\n        .map(b => b.toString(16).padStart(2, '0')).join('').substring(0, 12));\n    const blobKey = `findings--transcripts--email--${emailHash}`;\n    \n    console.log(`🔍 [${processingId}] Checking for duplicate email (current time: ${new Date(currentEmailTime).toISOString()})`);\n    \n    try {\n      const existingEmail = await blob.getJSON(blobKey);\n      if (existingEmail) {\n        const previousEmailTime = new Date(existingEmail.emailTimestamp).getTime();\n        const timeDifferenceMinutes = (currentEmailTime - previousEmailTime) / (1000 * 60);\n        \n        console.log(`🔍 [${processingId}] Found existing email from ${existingEmail.emailTimestamp}`);\n        console.log(`🔍 [${processingId}] Time difference: ${timeDifferenceMinutes.toFixed(2)} minutes`);\n        \n        if (timeDifferenceMinutes < 3) {\n          console.log(`⚠️ [${processingId}] Duplicate email detected! Already processed by ${existingEmail.processingId}`);\n          console.log(`⚠️ [${processingId}] Original email time: ${existingEmail.emailTimestamp}`);\n          return createSuccessResponse(processingId, `Duplicate email - sent only ${timeDifferenceMinutes.toFixed(2)} minutes after previous`);\n        } else {\n          console.log(`✅ [${processingId}] Email allowed - ${timeDifferenceMinutes.toFixed(2)} minutes > 3 minute threshold`);\n        }\n      }\n    } catch (blobError) {\n      console.log(`⚠️ [${processingId}] Blob lookup failed, continuing with processing: ${blobError.message}`);\n    }\n\n    // Store/update email processing metadata\n    try {\n      await blob.setJSON(blobKey, {\n        processed: true,\n        processingId,\n        timestamp: new Date().toISOString(),\n        emailTimestamp: new Date(currentEmailTime).toISOString(),\n        from: email.from,\n        subject: email.subject || 'No Subject',\n        attachmentCount: email.attachments?.length || 0,\n        emailHash\n      });\n      console.log(`✅ [${processingId}] Email marked as processing in blob storage`);\n    } catch (blobError) {\n      console.log(`⚠️ [${processingId}] Failed to store email metadata, continuing: ${blobError.message}`);\n    }\n\n    // Validate environment configuration\n    const databaseId = Deno.env.get('FINDINGS_TRANSCRIPTS_DB_ID');\n    if (!databaseId) {\n      console.error(`❌ [${processingId}] FINDINGS_TRANSCRIPTS_DB_ID environment variable not set`);\n      return createSuccessResponse(processingId, 'Configuration error logged');\n    }\n\n    // Validate email has attachments\n    if (!email.attachments || email.attachments.length === 0) {\n      console.log(`📧 [${processingId}] No attachments found in email`);\n      return createSuccessResponse(processingId, 'No attachments to process');\n    }\n\n    // Extract and validate PDF attachments\n    const pdfAttachments = extractPdfAttachments(email.attachments);\n    console.log(`📄 [${processingId}] Found ${pdfAttachments.length} PDF attachments`);\n\n    if (pdfAttachments.length === 0) {\n      console.log(`📧 [${processingId}] No PDF attachments found in email`);\n      return createSuccessResponse(processingId, 'No PDF attachments to process');\n    }\n\n    // Send immediate acknowledgment to user\n    console.log(`📧 [${processingId}] Sending acknowledgment email...`);\n    await sendProcessingAcknowledgment({\n      originalEmail: email,\n      processingId,\n      pdfCount: pdfAttachments.length\n    });\n\n    // Start async processing (fire and forget)\n    console.log(`🚀 [${processingId}] Starting async processing for ${pdfAttachments.length} PDFs...`);\n    processAttachmentsAsync(email.attachments, databaseId, processingId, email);\n    \n    // Return immediate success response to email service\n    return {\n      success: true,\n      message: `Started processing ${pdfAttachments.length} PDF attachments`,\n      processingId,\n      totalAttachments: email.attachments.length,\n      pdfAttachments: pdfAttachments.length,\n      timestamp: new Date().toISOString(),\n    };\n\n  } catch (error) {\n    console.error(`❌ [${processingId}] Email handler error:`, error);\n    // Always return success to prevent email service retries\n    return createSuccessResponse(processingId, 'Processing error logged', error.message);\n  }\n}\n\n/**\n * Extract PDF attachments from email attachments\n */\nfunction extractPdfAttachments(attachments: File[]): File[] {\n  console.log(`📎 Processing ${attachments.length} attachments:`);\n  attachments.forEach((file, index) => {\n    console.log(`   ${index + 1}. ${file.name} (${file.type}, ${file.size} bytes)`);\n  });\n\n  return attachments.filter(\n    file => file.name.toLowerCase().endsWith('.pdf') || \n            file.type.includes('pdf')\n  );\n}\n\n/**\n * Create standardized success response\n */\nfunction createSuccessResponse(processingId: string, message: string, error?: string) {\n  return {\n    success: true,\n    message,\n    processingId,\n    timestamp: new Date().toISOString(),\n    ...(error && { error })\n  };\n}\n\n/**\n * Process attachments asynchronously (business logic orchestration)\n * Handles communication with business logic and email service\n */\nasync function processAttachmentsAsync(\n  attachments: File[], \n  databaseId: string, \n  processingId: string, \n  originalEmail: Email\n) {\n  // Import email service functions here to avoid circular dependencies\n  const { sendProcessingCompletion, sendProcessingError } = await import(\"../services/email.service.ts\");\n  \n  try {\n    console.log(`🔄 [${processingId}] Starting async attachment processing...`);\n\n    // Call business logic controller\n    const result = await processEmailPdfAttachments({\n      attachments: attachments,\n      databaseId,\n      pdfPropertyName: 'PDF',\n    });\n\n    // Log processing results\n    console.log(`📊 [${processingId}] Async processing complete:`);\n    console.log(`   - Total attachments: ${result.totalAttachments}`);\n    console.log(`   - PDF attachments found: ${result.processedAttachments}`);\n    console.log(`   - Pages created: ${result.processedAttachments}`);\n    console.log(`   - Files uploaded: ${result.successfulUploads}`);\n    \n    if (result.errors.length > 0) {\n      console.log(`   - Errors: ${result.errors.length}`);\n      result.errors.forEach(error => console.log(`     • ${error}`));\n    }\n\n    // Send completion notification to user\n    await sendProcessingCompletion({\n      originalEmail,\n      processingId,\n      results: {\n        totalAttachments: result.totalAttachments,\n        processedAttachments: result.processedAttachments,\n        successfulUploads: result.successfulUploads,\n        errors: result.errors,\n        success: result.success\n      }\n    });\n\n    // Log final status\n    if (result.success) {\n      console.log(`✅ [${processingId}] All attachments processed successfully`);\n    } else {\n      console.log(`⚠️ [${processingId}] Processing completed with some errors`);\n    }\n\n  } catch (error) {\n    console.error(`❌ [${processingId}] Async processing error:`, error);\n    \n    // Send error notification to user\n    await sendProcessingError({\n      originalEmail,\n      processingId,\n      errorMessage: error.message\n    });\n  }\n}",[],[235,319,332],[235,319],{"_102":342,"_95":343,"_155":344,"_157":236,"_158":137,"_159":160,"_161":162,"_104":345,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_169":346,"_98":445},"d0f9da34-7951-11f0-8810-0224a6c84d84","routes","0198aa63-0ee2-725e-8c9f-d562adf43e32",["D",1755205275298],[347,369,392,422,429,437],{"_102":348,"_95":349,"_155":350,"_157":344,"_158":137,"_159":160,"_161":162,"_104":351,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_169":352,"_98":368},"d0f9c0b2-7951-11f0-8810-0224a6c84d84","api","0198aa63-0ee2-725e-8c9f-abba08232a49",["D",1755205275298],[353,360],{"_102":354,"_95":247,"_155":355,"_157":350,"_158":137,"_159":160,"_161":162,"_104":356,"_99":32,"_164":177,"_166":357,"_167":22,"_168":-5,"_169":358,"_98":359},"d0f9bd24-7951-11f0-8810-0224a6c84d84","0198aa63-0ee2-725e-8c9f-a1731ad70c90",["D",1755205275298],"# API Routes\n\nJSON API endpoints for frontend/backend communication.",[],[235,343,349,247],{"_102":361,"_95":362,"_155":363,"_157":350,"_158":137,"_159":160,"_161":162,"_104":364,"_99":32,"_164":260,"_166":365,"_167":22,"_168":-5,"_169":366,"_98":367},"d0f9bf9a-7951-11f0-8810-0224a6c84d84","_api.routes.ts","0198aa63-0ee2-725e-8c9f-a4a44f77a838",["D",1755205275298],"import { Hono } from \"npm:hono@3.12.12\";\nimport { getHealthStatus } from \"../../controllers/health.controller.ts\";\n\nconst app = new Hono();\n\napp.get(\"/health\", async (c) => {\n  const healthStatus = await getHealthStatus();\n  return c.json(healthStatus);\n});\n\nexport default app;",[],[235,343,349,362],[235,343,349],{"_102":370,"_95":371,"_155":372,"_157":344,"_158":137,"_159":160,"_161":162,"_104":373,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_169":374,"_98":391},"d0f9c562-7951-11f0-8810-0224a6c84d84","tasks","0198aa63-0ee2-725e-8c9f-b6bbe70d9965",["D",1755205275298],[375,383],{"_102":376,"_95":247,"_155":377,"_157":372,"_158":137,"_159":268,"_161":162,"_104":378,"_99":379,"_164":177,"_166":380,"_167":22,"_168":-5,"_169":381,"_98":382},"8ab51650-840b-11f0-a09a-0224a6c84d84","0198aa63-0ee2-725e-8c9f-ae02d77fc854",["D",1756384555490],4,"# Task Routes\n\nWebhook handlers for Notion integrations.\n\n## Endpoints\n\n- `POST /tasks/blob` - Process PDF extraction and blob storage for Notion pages\n- `POST /tasks/dupe` - Duplicate Notion page to final reports database with content filtering\n- `POST /tasks/convert` - Convert page to database based on H2 sections\n- `POST /tasks/draft` - Duplicate page with filtered child database (excludes specific analysis properties)\n- `POST /tasks/notion-webhook` - General Notion webhook handler\n- `POST /tasks/test` - Test endpoint for webhook authentication\n- `GET /tasks/debug-webhook` - Debug endpoint to check webhook configuration\n\n## Architecture\n\nRoutes handle HTTP request/response only. Business logic is delegated to:\n\n- `tasks.controller.ts` - Orchestrates PDF extraction, blob storage, page duplication, and Notion updates\n- `notion.service.ts` - Handles Notion API calls\n- `pdfExtractor.ts` - Handles PDF text extraction\n\n## Page Duplication (`/tasks/dupe`)\n\nThe `/tasks/dupe` endpoint duplicates a Notion page from any database to the final reports database (`FINDINGS_REPORTS_DRAFT_DB_ID`) with content filtering:\n\n### Content Filtering\n\n- **Removes**: Callout blocks, code blocks, divider blocks, blocks containing \"Evidence summary\", blocks containing \"Methodology\", blocks containing \"Executive summary\"\n- **Preserves**: All other content blocks in original order\n- **Source**: Original page remains unchanged\n\n### Webhook Payload\n\nExpects Notion webhook with page ID in `data.id`:\n\n```json\n{\n  \"data\": {\n    \"id\": \"page-id-here\"\n  }\n}\n```\n\n### Response Format\n\n```json\n{\n  \"success\": true,\n  \"sourcePageId\": \"source-page-id\",\n  \"targetPageId\": \"new-page-id\",\n  \"blocksCopied\": 15,\n  \"blocksFiltered\": 3,\n  \"timestamp\": \"2025-08-15T22:25:47.000Z\"\n}\n```\n\n## Page Duplication with Filtered Database (`/tasks/draft`)\n\nThe `/tasks/draft` endpoint duplicates a Notion page and its child database to the draft reports database (`FINDINGS_REPORTS_DRAFT_DB_ID`) with selective property filtering:\n\n### Child Database Detection\n\n- Uses \"Results Database ID\" property from source page to identify child database\n- If no child database ID found, proceeds with page duplication only\n\n### Property Filtering\n\n**Excluded from child database**:\n- Confidence Score\n- Consensus Level  \n- Evidence Summary\n- Methodology\n\n**Preserved in child database**:\n- Name (title)\n- Status\n- Category\n- Priority\n- All other properties\n\n### Content Processing\n\n- **Page Content**: Copies all blocks except child_database blocks\n- **Child Database**: Creates new filtered database with all pages duplicated\n- **Name Property**: Replaces \"(raw)\" with \"(draft)\" or appends \"(draft)\"\n\n### Response Format\n\n```json\n{\n  \"success\": true,\n  \"sourcePageId\": \"source-page-id\",\n  \"targetPageId\": \"new-page-id\", \n  \"childDatabaseId\": \"source-child-db-id\",\n  \"targetChildDatabaseId\": \"new-child-db-id\",\n  \"childPagesProcessed\": 25,\n  \"childPagesSucceeded\": 25,\n  \"timestamp\": \"2025-08-15T22:25:47.000Z\"\n}\n```\n",[],[235,343,371,247],{"_102":384,"_95":385,"_155":386,"_157":372,"_158":137,"_159":268,"_161":162,"_104":387,"_99":379,"_164":260,"_166":388,"_167":22,"_168":-5,"_169":389,"_98":390},"8ab51bfa-840b-11f0-a09a-0224a6c84d84","_tasks.routes.ts","0198aa63-0ee2-725e-8c9f-b3edc7440c83",["D",1756384555490],"import { Hono } from \"npm:hono@3.12.12\";\nimport { processBlobExtraction, duplicatePageToFinalReports, convertPageToDatabase, duplicatePageWithFilteredDatabase } from \"../../controllers/tasks.controller.ts\";\nimport { extractPageIdFromWebhook } from \"../../utils/notion-properties.utils.ts\";\n\nconst app = new Hono();\n\n// Debug endpoint to check webhook configuration\napp.get(\"/debug-webhook\", (c) => {\n  const secret = Deno.env.get(\"NOTION_WEBHOOK_SECRET\");\n  return c.json({\n    hasSecret: !!secret,\n    secretLength: secret?.length || 0,\n  });\n});\n\n// Test endpoint for webhook authentication\napp.post(\"/test\", async (c) => {\n  // At this point, webhook auth has already passed\n  return c.json({ \n    success: true, \n    message: \"Webhook authentication successful\",\n    timestamp: new Date().toISOString()\n  });\n});\n\n// Main Notion webhook endpoint\napp.post(\"/notion-webhook\", async (c) => {\n  // At this point, webhook auth has already passed\n  const body = await c.req.json();\n  console.log(\"Notion webhook received:\", body);\n\n  // Process the webhook...\n  // TODO: Add webhook processing logic here\n\n  return c.json({ success: true });\n});\n\n// PDF text extraction and blob storage endpoint\napp.post(\"/blob\", async (c) => {\n  try {\n    // At this point, webhook auth has already passed\n    const body = await c.req.json();\n    console.log(\"📥 Blob processing webhook received:\", body);\n\n    // Extract page ID from webhook payload\n    const pageId = extractPageIdFromWebhook(body);\n    if (!pageId) {\n      return c.json({ error: \"Page ID is required\" }, 400);\n    }\n\n    // Process the blob extraction using controller\n    const result = await processBlobExtraction({ pageId });\n\n    // Return appropriate HTTP status based on result\n    if (result.success) {\n      return c.json(result);\n    } else {\n      const statusCode = result.error?.includes(\"No PDF\") || result.error?.includes(\"No text\") ? 400 : 500;\n      return c.json(result, statusCode);\n    }\n\n  } catch (error) {\n    console.error(\"❌ Error in blob endpoint:\", error);\n    return c.json({ \n      error: \"Internal server error\", \n      details: error.message,\n      timestamp: new Date().toISOString()\n    }, 500);\n  }\n});\n\n// Page duplication endpoint\napp.post(\"/dupe\", async (c) => {\n  try {\n    // At this point, webhook auth has already passed\n    const body = await c.req.json();\n    console.log(\"📥 Page duplication webhook received:\", body);\n\n    // Extract page ID from webhook payload\n    const pageId = extractPageIdFromWebhook(body);\n    if (!pageId) {\n      return c.json({ error: \"Page ID is required\" }, 400);\n    }\n\n    // Process the page duplication using controller\n    const result = await duplicatePageToFinalReports({ pageId });\n\n    // Return appropriate HTTP status based on result\n    if (result.success) {\n      return c.json(result);\n    } else {\n      return c.json(result, 500);\n    }\n\n  } catch (error) {\n    console.error(\"❌ Error in dupe endpoint:\", error);\n    return c.json({ \n      error: \"Internal server error\", \n      details: error.message,\n      timestamp: new Date().toISOString()\n    }, 500);\n  }\n});\n\n// Page conversion endpoint\napp.post(\"/convert\", async (c) => {\n  try {\n    // At this point, webhook auth has already passed\n    const body = await c.req.json();\n    console.log(\"📥 Page conversion webhook received:\", body);\n\n    // Extract page ID from webhook payload\n    const pageId = extractPageIdFromWebhook(body);\n    if (!pageId) {\n      return c.json({ error: \"Page ID is required\" }, 400);\n    }\n\n    // Process the page conversion using controller\n    const result = await convertPageToDatabase({ pageId });\n\n    // Return appropriate HTTP status based on result\n    if (result.success) {\n      return c.json(result);\n    } else {\n      return c.json(result, 500);\n    }\n\n  } catch (error) {\n    console.error(\"❌ Error in convert endpoint:\", error);\n    return c.json({ \n      error: \"Internal server error\", \n      details: error.message,\n      timestamp: new Date().toISOString()\n    }, 500);\n  }\n});\n\n// Page duplication with filtered database endpoint\napp.post(\"/draft\", async (c) => {\n  try {\n    // At this point, webhook auth has already passed\n    const body = await c.req.json();\n    console.log(\"📥 Draft duplication webhook received:\", body);\n\n    // Extract page ID from webhook payload\n    const pageId = extractPageIdFromWebhook(body);\n    if (!pageId) {\n      return c.json({ error: \"Page ID is required\" }, 400);\n    }\n\n    // Process the page duplication with filtered database using controller\n    const result = await duplicatePageWithFilteredDatabase({ pageId });\n\n    // Return appropriate HTTP status based on result\n    if (result.success) {\n      return c.json(result);\n    } else {\n      return c.json(result, 500);\n    }\n\n  } catch (error) {\n    console.error(\"❌ Error in draft endpoint:\", error);\n    return c.json({ \n      error: \"Internal server error\", \n      details: error.message,\n      timestamp: new Date().toISOString()\n    }, 500);\n  }\n});\n\n// Catch-all handler for undefined webhook endpoints\n// This must be the LAST route defined to catch unmatched paths\napp.post(\"*\", (c) => {\n  const path = c.req.path;\n  const method = c.req.method;\n  const returnObj = {\n    error: \"Endpoint not found\",\n    path: path,\n    method: method,\n    availableEndpoints: [\n      \"GET /tasks/debug-webhook\",\n      \"POST /tasks/test\",\n      \"POST /tasks/notion-webhook\",\n      \"POST /tasks/blob\",\n      \"POST /tasks/dupe\",\n      \"POST /tasks/convert\",\n      \"POST /tasks/draft\",\n    ],\n  };\n  console.log(returnObj);\n\n  return c.json(returnObj, 404);\n});\n\nexport default app;",[],[235,343,371,385],[235,343,371],{"_102":393,"_95":394,"_155":395,"_157":344,"_158":137,"_159":160,"_161":162,"_104":396,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_169":397,"_98":421},"d0f9d336-7951-11f0-8810-0224a6c84d84","views","0198aa63-0ee2-725e-8c9f-c76212a512de",["D",1755205275298],[398,405,413],{"_102":399,"_95":247,"_155":400,"_157":395,"_158":137,"_159":160,"_161":162,"_104":401,"_99":32,"_164":177,"_166":402,"_167":22,"_168":-5,"_169":403,"_98":404},"d0f9c97c-7951-11f0-8810-0224a6c84d84","0198aa63-0ee2-725e-8c9f-b8ca6925a52f",["D",1755205275298],"# View Routes\n\nServer-side rendered views for frontend.",[],[235,343,394,247],{"_102":406,"_95":407,"_155":408,"_157":395,"_158":137,"_159":160,"_161":162,"_104":409,"_99":32,"_164":260,"_166":410,"_167":22,"_168":-5,"_169":411,"_98":412},"d0f9cb02-7951-11f0-8810-0224a6c84d84","_views.routes.ts","0198aa63-0ee2-725e-8c9f-bce257a069f1",["D",1755205275298],"import { Hono } from \"npm:hono@3.12.12\";\n\nconst app = new Hono();\n\n// TODO: Add routes here\n\nexport default app;",[],[235,343,394,407],{"_102":414,"_95":415,"_155":416,"_157":395,"_158":137,"_159":160,"_161":162,"_104":417,"_99":32,"_164":260,"_166":418,"_167":22,"_168":-5,"_169":419,"_98":420},"d0f9cc06-7951-11f0-8810-0224a6c84d84","dashboard.tsx","0198aa63-0ee2-725e-8c9f-c128742e671b",["D",1755205275298],"import { getAbsoluteUrl } from \"../../utils/url.ts\";\nimport { getHealthStatus } from \"../../controllers/health.controller.ts\";\n\nexport default async (c) => {\n  const userEmail = c.get(\"userEmail\");\n  const healthData = await getHealthStatus();\n  const healthUrl = getAbsoluteUrl(c, \"/api/health\");\n  \n  // Check webhook configuration\n  const webhookSecret = Deno.env.get(\"NOTION_WEBHOOK_SECRET\");\n  const webhookConfigured = !!webhookSecret;\n  const webhookEndpoint = getAbsoluteUrl(c, '/tasks/notion-webhook');\n\n  return c.html(`\n    \n    \n    \n      Findings Dashboard\n      \n    \n    \n      \n        \n          Findings Dashboard
\n          \n            Welcome, ${userEmail}\n            Logout\n          
\n        
\n        \n        \n          \n            Welcome to your Findings Dashboard
\n            This is your central hub for managing research findings, transcripts, questions, and reports.
\n          
\n\n          \n            \n              System Health Status\n              \n                ${healthData.status}\n              \n            
\n            Current system status and database connectivity:
\n            \n            ${JSON.stringify(healthData, null, 2)}
\n            \n            \n              View Raw Health Endpoint →\n            \n\n            \n              Webhook Endpoint
\n\n              \n                ${webhookEndpoint}
\n              
\n\n              \n                To test: Get the NOTION_WEBHOOK_SECRET value from your vars.env file (provided by your administrator)
\n                \n                  \n                  \n                
\n                Ready to test webhook authentication...
\n              
\n            
\n          
\n        
\n      
\n\n      \n    \n    \n  `);\n};",[],[235,343,394,415],[235,343,394],{"_102":423,"_95":247,"_155":424,"_157":344,"_158":137,"_159":160,"_161":162,"_104":425,"_99":32,"_164":177,"_166":426,"_167":22,"_168":-5,"_169":427,"_98":428},"d0f9d458-7951-11f0-8810-0224a6c84d84","0198aa63-0ee2-725e-8c9f-cbd69df307e2",["D",1755205275298],"# Routes\n\nHTTP route handlers organized by functionality. Each subdirectory has a `_[directory_name].routes.ts` file that mounts routes for that section.",[],[235,343,247],{"_102":430,"_95":431,"_155":432,"_157":344,"_158":137,"_159":160,"_161":162,"_104":433,"_99":32,"_164":260,"_166":434,"_167":22,"_168":-5,"_169":435,"_98":436},"d0f9d566-7951-11f0-8810-0224a6c84d84","authCheck.ts","0198aa63-0ee2-725e-8c9f-cdfa4503b01e",["D",1755205275298],"// Public routes that don't require authentication\nconst PUBLIC_ROUTES = [\n  \"/api/health\", // Health check endpoint should be public\n  // Add other public routes here as needed\n  // '/demo/:id',  // Example: public demo viewing\n  // '/api/public', // Example: public API endpoints\n];\n\n// Helper function to check if a route is public\nfunction isPublicRoute(path: string): boolean {\n  return PUBLIC_ROUTES.some((route) => {\n    // Handle exact matches\n    if (route === path) return true;\n\n    // Handle wildcard patterns (basic implementation)\n    if (route.includes(\"*\")) {\n      const pattern = route.replace(\"*\", \".*\");\n      return new RegExp(`^${pattern}$`).test(path);\n    }\n\n    // Handle route parameters (basic implementation)\n    if (route.includes(\":\")) {\n      const pattern = route.replace(/:[^/]+/g, \"[^/]+\");\n      return new RegExp(`^${pattern}$`).test(path);\n    }\n\n    return false;\n  });\n}\n\nexport default async (c, next) => {\n  // Allow public routes to bypass authentication\n  if (isPublicRoute(c.req.path)) {\n    await next();\n    return;\n  }\n\n  const email = c.req.header(\"X-LastLogin-Email\");\n\n  if (!email) {\n    // Show login page if not authenticated\n    return c.html(`\n      \n      \n      \n        Findings - Login Required\n        \n      \n      \n        \n          Sign in to Findings
\n          Sign in to access your dashboard and manage your research findings.
\n          \n            Loading login button...
\n          
\n          \n            via LastLogin\n          
\n        
\n\n        \n      \n      \n    `);\n  }\n\n  // If authenticated, store email in context for use by routes\n  c.set(\"userEmail\", email);\n\n  // Continue to the next middleware/route\n  await next();\n};",[],[235,343,431],{"_102":438,"_95":439,"_155":440,"_157":344,"_158":137,"_159":160,"_161":162,"_104":441,"_99":32,"_164":260,"_166":442,"_167":22,"_168":-5,"_169":443,"_98":444},"d0f9d8ae-7951-11f0-8810-0224a6c84d84","webhookAuthCheck.ts","0198aa63-0ee2-725e-8c9f-d0c7bb4f3135",["D",1755205275298],"export default async (c, next) => {\n  // Allow GET requests for debug endpoints to bypass webhook auth\n  if (c.req.method === \"GET\") {\n    await next();\n    return;\n  }\n\n  const apiKey = c.req.header(\"X-API-KEY\");\n  const expectedSecret = Deno.env.get(\"NOTION_WEBHOOK_SECRET\");\n\n  // Check if webhook secret is configured\n  if (!expectedSecret) {\n    console.log(\"Webhook authentication failed: NOTION_WEBHOOK_SECRET not configured\");\n    return c.json({ error: \"Authentication required\" }, 401);\n  }\n\n  // Check if API key is provided\n  if (!apiKey) {\n    console.log(\"Webhook authentication failed: No X-API-KEY header provided\");\n    return c.json({ error: \"Authentication required\" }, 401);\n  }\n\n  // Use constant-time comparison to prevent timing attacks\n  const expectedBuffer = new TextEncoder().encode(expectedSecret);\n  const providedBuffer = new TextEncoder().encode(apiKey);\n\n  // Ensure buffers are the same length for constant-time comparison\n  if (expectedBuffer.length !== providedBuffer.length) {\n    console.log(\"Webhook authentication failed: Invalid API key length\");\n    return c.json({ error: \"Authentication failed\" }, 403);\n  }\n\n  // Constant-time comparison\n  let isValid = true;\n  for (let i = 0; i < expectedBuffer.length; i++) {\n    if (expectedBuffer[i] !== providedBuffer[i]) {\n      isValid = false;\n    }\n  }\n\n  if (!isValid) {\n    console.log(\"Webhook authentication failed: Invalid API key\");\n    return c.json({ error: \"Authentication failed\" }, 403);\n  }\n\n  // Authentication successful, proceed to route handler\n  await next();\n};",[],[235,343,439],[235,343],{"_102":447,"_95":448,"_155":449,"_157":236,"_158":137,"_159":160,"_161":162,"_104":450,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_169":451,"_98":485},"d0f9dd9a-7951-11f0-8810-0224a6c84d84","services","0198aa63-0ee2-725e-8c9f-e39dc7a1bd77",["D",1755205275298],[452,459,468,476],{"_102":453,"_95":247,"_155":454,"_157":449,"_158":137,"_159":218,"_161":162,"_104":455,"_99":303,"_164":177,"_166":456,"_167":22,"_168":-5,"_169":457,"_98":458},"fdcaab68-841b-11f0-bd95-0224a6c84d84","0198aa63-0ee2-725e-8c9f-da1f9466971b",["D",1756391620516],"# Services\n\nService layer with proper separation of concerns. Each service handles a specific domain:\n\n## Architecture Principles\n\n### Separation of Concerns\n- **Email Service** (`email.service.ts`): Email communication and template formatting\n- **Notion Service** (`notion.service.ts`): Notion API integration for database and file operations  \n- **OpenAI Service** (`openai.service.ts`): AI/ML processing and analysis\n\n### Service Boundaries\nServices should:\n- ✅ Have a single responsibility\n- ✅ Be reusable across different controllers\n- ✅ Handle their own error cases gracefully\n- ✅ Return consistent response formats\n- ❌ Not call controllers or handlers directly\n- ❌ Not mix presentation logic with business logic\n\n## Available Services\n\n### Email Service (`email.service.ts`)\n**Purpose**: Centralized email communication with template management\n\n**Key Functions**:\n- `sendProcessingAcknowledgment()` - Immediate user feedback\n- `sendProcessingCompletion()` - Final status notification  \n- `sendProcessingError()` - Error notifications\n\n**Features**:\n- Text and HTML email templates\n- Consistent styling and branding\n- Graceful error handling (email failures don't break processing)\n\n### Notion Service (`notion.service.ts`)\n**Purpose**: Notion API integration for database queries and page updates\n\n**Key Functions**:\n- **Generic functions**: `updatePageProperties()` for any property updates\n- **Content management**: Page content creation, updates, and deletion\n- **Database operations**: Query databases and retrieve pages\n- **File Upload API**: 3-step file upload process implementation\n  - `createNotionFileUpload()` - Step 1: Create upload object\n  - `uploadFileToNotionUrl()` - Step 2: Upload file contents via FormData\n  - `attachFileUploadToPage()` - Step 3: Attach file to page with \"file_upload\" type\n\n### OpenAI Service (`openai.service.ts`)\n**Purpose**: OpenAI API integration with two-stage consensus analysis\n\n**Key Functions**:\n- **Stage 1**: Individual transcript analysis using `gpt-4o-mini` for cost efficiency\n- **Stage 2**: Consensus synthesis using `gpt-4o` for high-quality reasoning\n\n**Features**:\n- Structured JSON output, confidence scoring, disagreement tracking, evidence tables\n- Configurable parallel processing with rate limiting\n- Graceful error handling with fallback consensus calculation\n\n## Usage Examples\n\n### Email Service\n```typescript\nimport { sendProcessingAcknowledgment } from \"../services/email.service.ts\";\n\nawait sendProcessingAcknowledgment({\n  originalEmail: email,\n  processingId: \"abc123\",\n  pdfCount: 5\n});\n```\n\n### Testing\nUse `/test-async-email-processing?email=your@email.com&count=3` to test the complete workflow.",[],[235,448,247],{"_102":460,"_95":461,"_155":462,"_157":449,"_158":137,"_159":463,"_161":162,"_104":464,"_99":303,"_164":260,"_166":465,"_167":22,"_168":-5,"_169":466,"_98":467},"1cdcdbac-8444-11f0-a238-0224a6c84d84","email.service.ts","fdca8e8a-841b-11f0-bd95-0224a6c84d84",42,["D",1756408852517],"/**\n * Email Service\n * Handles all email communication and template formatting\n * Separated from business logic for better maintainability\n */\n\nimport { email } from \"https://esm.town/v/std/email\";\nimport type { Email } from \"../types/email.types.ts\";\n\nexport interface EmailNotificationData {\n  originalEmail: Email;\n  processingId: string;\n  pdfCount?: number;\n  estimatedMinutes?: number;\n  results?: {\n    totalAttachments: number;\n    processedAttachments: number;\n    successfulUploads: number;\n    errors: string[];\n    success: boolean;\n  };\n  errorMessage?: string;\n}\n\n/**\n * Send immediate acknowledgment email to sender\n */\nexport async function sendProcessingAcknowledgment(data: EmailNotificationData): Promise {\n  try {\n    const { originalEmail, processingId, pdfCount = 0 } = data;\n    const estimatedMinutes = Math.ceil(pdfCount * 0.5);\n    const maxMinutes = Math.ceil(pdfCount * 1.5);\n    \n    const toAddress = Array.isArray(originalEmail.to) \n      ? originalEmail.to[0] \n      : originalEmail.to || 'findings@valtown.email';\n    const emailPrefix = toAddress.split('@')[0] + '@';\n    const subject = `${emailPrefix} - transcript processing started`;\n    \n    const textContent = createAcknowledgmentTextTemplate({\n      originalSubject: originalEmail.subject || 'No Subject',\n      pdfCount,\n      processingId,\n      estimatedMinutes,\n      maxMinutes\n    });\n\n    await email({\n      to: originalEmail.from,\n      subject: subject,\n      text: textContent\n    });\n\n    console.log(`📧 [${processingId}] Acknowledgment email sent to ${originalEmail.from}`);\n  } catch (error) {\n    console.error(`❌ Failed to send acknowledgment email:`, error);\n    // Don't throw - email failures shouldn't break processing\n  }\n}\n\n/**\n * Send processing completion status email to sender\n */\nexport async function sendProcessingCompletion(data: EmailNotificationData): Promise {\n  try {\n    const { originalEmail, processingId, results } = data;\n    \n    if (!results) {\n      throw new Error('Results data required for completion email');\n    }\n    \n    const toAddress = Array.isArray(originalEmail.to) \n      ? originalEmail.to[0] \n      : originalEmail.to || 'findings@valtown.email';\n    const emailPrefix = toAddress.split('@')[0] + '@';\n    const isSuccess = results.success && results.errors.length === 0;\n    const subject = `${emailPrefix} - transcript processing ${isSuccess ? 'complete' : 'complete with errors'}`;\n    \n    const textContent = createCompletionTextTemplate({\n      originalSubject: originalEmail.subject || 'No Subject',\n      processingId,\n      results,\n      isSuccess\n    });\n\n    await email({\n      to: originalEmail.from,\n      subject: subject,\n      text: textContent\n    });\n\n    console.log(`📧 [${processingId}] Completion email sent to ${originalEmail.from}`);\n  } catch (error) {\n    console.error(`❌ Failed to send completion email:`, error);\n    // Don't throw - email failures shouldn't break processing\n  }\n}\n\n/**\n * Send error notification email to sender\n */\nexport async function sendProcessingError(data: EmailNotificationData): Promise {\n  try {\n    const { originalEmail, processingId, errorMessage } = data;\n    \n    if (!errorMessage) {\n      throw new Error('Error message required for error notification email');\n    }\n    \n    const toAddress = Array.isArray(originalEmail.to) \n      ? originalEmail.to[0] \n      : originalEmail.to || 'findings@valtown.email';\n    const emailPrefix = toAddress.split('@')[0] + '@';\n    const subject = `${emailPrefix} - transcript processing failed`;\n    \n    const textContent = createErrorTextTemplate({\n      originalSubject: originalEmail.subject || 'No Subject',\n      processingId,\n      errorMessage\n    });\n\n    await email({\n      to: originalEmail.from,\n      subject: subject,\n      text: textContent\n    });\n\n    console.log(`📧 [${processingId}] Error notification sent to ${originalEmail.from}`);\n  } catch (error) {\n    console.error(`❌ Failed to send error notification:`, error);\n    // Don't throw - email failures shouldn't break processing\n  }\n}\n\n// ============================================================================\n// EMAIL TEMPLATES\n// ============================================================================\n\ninterface AcknowledgmentTemplateData {\n  originalSubject: string;\n  pdfCount: number;\n  processingId: string;\n  estimatedMinutes: number;\n  maxMinutes: number;\n}\n\nfunction createAcknowledgmentTextTemplate(data: AcknowledgmentTemplateData): string {\n  return `Original subject line: ${data.originalSubject}\nPDF attachments: ${data.pdfCount}\n\nYour PDF files are being processed and uploaded to the database in Notion. You'll receive another email when processing is complete.\n\nProcess ID: ${data.processingId}`;\n}\n\ninterface CompletionTemplateData {\n  originalSubject: string;\n  processingId: string;\n  results: {\n    totalAttachments: number;\n    processedAttachments: number;\n    successfulUploads: number;\n    errors: string[];\n  };\n  isSuccess: boolean;\n}\n\nfunction createCompletionTextTemplate(data: CompletionTemplateData): string {\n  let text = `Original subject line: ${data.originalSubject}\nProcessing results: ${data.isSuccess ? 'Success' : 'Completed with errors'}\n\nTotal attachments: ${data.results.totalAttachments}\nPDF files processed: ${data.results.processedAttachments}\nPages created in Notion: ${data.results.processedAttachments}\nFiles uploaded successfully: ${data.results.successfulUploads}`;\n\n  if (data.results.errors.length > 0) {\n    text += `\\n\\nErrors encountered:\\n${data.results.errors.map(error => `- ${error}`).join('\\n')}`;\n  }\n\n  text += `\\n\\nYour files are now available in the Notion database.\n\nProcess ID: ${data.processingId}\nProcessing completed: ${new Date().toLocaleString()}`;\n\n  return text;\n}\n\ninterface ErrorTemplateData {\n  originalSubject: string;\n  processingId: string;\n  errorMessage: string;\n}\n\nfunction createErrorTextTemplate(data: ErrorTemplateData): string {\n  return `❌ Processing failed for your email: ${data.originalSubject}\n\n🚨 Error Details:\n${data.errorMessage}\n\n🆔 Processing ID: ${data.processingId} (for support reference)\n\nPlease try sending your email again, or contact support if the issue persists.\n\nError occurred at: ${new Date().toLocaleString()}`;\n}",[],[235,448,461],{"_102":469,"_95":470,"_155":471,"_157":449,"_158":137,"_159":218,"_161":162,"_104":472,"_99":227,"_164":260,"_166":473,"_167":22,"_168":-5,"_169":474,"_98":475},"fdcaae06-841b-11f0-bd95-0224a6c84d84","notion.service.ts","0198aa63-0ee2-725e-8c9f-deb04ed4654f",["D",1756391620516],"import { Client } from \"npm:@notionhq/client\";\n\nconst notion = new Client({ auth: Deno.env.get(\"NOTION_API_KEY\") });\n\nexport async function getDatabases() {\n  try {\n    const response = await notion.search({\n      filter: {\n        property: \"object\",\n        value: \"database\",\n      },\n    });\n    return {\n      success: true,\n      data: response,\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\nexport async function getDatabaseById(databaseId: string) {\n  try {\n    const response = await notion.databases.retrieve({\n      database_id: databaseId,\n    });\n    return {\n      success: true,\n      data: response,\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\nexport async function getPageProperties(pageId: string) {\n  try {\n    const response = await notion.pages.retrieve({\n      page_id: pageId,\n    });\n    return {\n      success: true,\n      data: response,\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\nexport async function updatePageStatus(pageId: string, status: string, blobKey?: string) {\n  try {\n    const properties: any = {\n      \"Status\": {\n        select: {\n          name: status\n        }\n      }\n    };\n\n    // Add blob key if provided\n    if (blobKey) {\n      properties[\"Blob key\"] = {\n        rich_text: [\n          {\n            text: {\n              content: blobKey\n            }\n          }\n        ]\n      };\n    }\n\n    const response = await notion.pages.update({\n      page_id: pageId,\n      properties: properties\n    });\n    \n    return {\n      success: true,\n      data: response,\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\nexport async function queryDatabasePages(databaseId: string, filter?: any) {\n  try {\n    const response = await notion.databases.query({\n      database_id: databaseId,\n      filter: filter\n    });\n    \n    return {\n      success: true,\n      data: response,\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\nexport async function getDatabasePages(databaseId: string) {\n  try {\n    const response = await notion.databases.query({\n      database_id: databaseId\n    });\n    \n    return {\n      success: true,\n      data: response,\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\nexport async function updatePageContent(pageId: string, blocks: any[]) {\n  try {\n    // First, get existing blocks to clear them\n    const existingBlocks = await notion.blocks.children.list({\n      block_id: pageId\n    });\n\n    // Delete existing blocks\n    for (const block of existingBlocks.results) {\n      await notion.blocks.delete({\n        block_id: block.id\n      });\n    }\n\n    // Add new blocks\n    const response = await notion.blocks.children.append({\n      block_id: pageId,\n      children: blocks\n    });\n    \n    return {\n      success: true,\n      data: response,\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\n/**\n * Delete all block children from a page (clean slate)\n */\nexport async function deletePageContent(pageId: string) {\n  try {\n    const existingBlocks = await notion.blocks.children.list({\n      block_id: pageId\n    });\n\n    // Delete all existing blocks\n    for (const block of existingBlocks.results) {\n      await notion.blocks.delete({\n        block_id: block.id\n      });\n    }\n    \n    return {\n      success: true,\n      data: { deletedBlocks: existingBlocks.results.length },\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\n/**\n * Append new blocks to existing page content (incremental updates)\n */\nexport async function appendPageContent(pageId: string, blocks: any[]) {\n  try {\n    const response = await notion.blocks.children.append({\n      block_id: pageId,\n      children: blocks\n    });\n    \n    return {\n      success: true,\n      data: response,\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\n/**\n * Update page properties (generic function for any property updates)\n */\nexport async function updatePageProperties(pageId: string, properties: any) {\n  try {\n    const response = await notion.pages.update({\n      page_id: pageId,\n      properties: properties\n    });\n    \n    return {\n      success: true,\n      data: response,\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\n/**\n * Get all content blocks from a page (with pagination)\n */\nexport async function getPageBlocks(pageId: string) {\n  try {\n    let allBlocks: any[] = [];\n    let hasMore = true;\n    let nextCursor: string | undefined = undefined;\n    \n    while (hasMore) {\n      const response = await notion.blocks.children.list({\n        block_id: pageId,\n        page_size: 100,\n        start_cursor: nextCursor\n      });\n      \n      allBlocks = allBlocks.concat(response.results);\n      hasMore = response.has_more;\n      nextCursor = response.next_cursor || undefined;\n      \n      console.log(`📄 Retrieved ${response.results.length} blocks (total: ${allBlocks.length}, hasMore: ${hasMore})`);\n    }\n    \n    return {\n      success: true,\n      data: { results: allBlocks },\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\n/**\n * Create a new page in a database\n */\nexport async function createPageInDatabase(databaseId: string, properties: any) {\n  try {\n    const response = await notion.pages.create({\n      parent: {\n        database_id: databaseId\n      },\n      properties: properties\n    });\n    \n    return {\n      success: true,\n      data: response,\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\n/**\n * Create a new page in a database with content\n */\nexport async function createPageInDatabaseWithContent(databaseId: string, properties: any, children: any[]) {\n  try {\n    const response = await notion.pages.create({\n      parent: {\n        database_id: databaseId\n      },\n      properties: properties,\n      children: children\n    });\n    \n    return {\n      success: true,\n      data: response,\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\n/**\n * Create an inline database within a page\n */\nexport async function createInlineDatabase(pageId: string, title: string, properties: any) {\n  try {\n    const response = await notion.databases.create({\n      parent: {\n        type: \"page_id\",\n        page_id: pageId,\n      },\n      is_inline: true,\n      title: [\n        {\n          type: \"text\",\n          text: {\n            content: title,\n          },\n        },\n      ],\n      properties: properties,\n    });\n    \n    return {\n      success: true,\n      data: {\n        databaseId: response.id,\n        database: response\n      },\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\n/**\n * Copy blocks to a target page (with batching and retry logic)\n */\nexport async function copyBlocksToPage(targetPageId: string, blocks: any[]) {\n  try {\n    // Log the blocks we're trying to copy for debugging\n    console.log(`🔧 Attempting to copy ${blocks.length} blocks to page ${targetPageId}`);\n    console.log(`🔧 Block types:`, blocks.map(b => b.type));\n    \n    // Notion API limit is 100 blocks per request, so we need to batch\n    const BATCH_SIZE = 50; // Reduced from 100 to be more conservative\n    const batches = [];\n    \n    for (let i = 0; i < blocks.length; i += BATCH_SIZE) {\n      batches.push(blocks.slice(i, i + BATCH_SIZE));\n    }\n    \n    console.log(`🔧 Splitting into ${batches.length} batches of up to ${BATCH_SIZE} blocks each`);\n    \n    let totalCopied = 0;\n    let failedBatches = 0;\n    \n    for (let i = 0; i < batches.length; i++) {\n      const batch = batches[i];\n      console.log(`🔧 Copying batch ${i + 1}/${batches.length} (${batch.length} blocks)`);\n      \n      // Retry logic for each batch\n      let batchSuccess = false;\n      let retryCount = 0;\n      const maxRetries = 2;\n      \n      while (!batchSuccess && retryCount <= maxRetries) {\n        try {\n          const response = await notion.blocks.children.append({\n            block_id: targetPageId,\n            children: batch\n          });\n          \n          batchSuccess = true;\n          totalCopied += batch.length;\n          console.log(`✅ Batch ${i + 1} copied successfully (${totalCopied}/${blocks.length} total)`);\n          \n        } catch (batchError) {\n          retryCount++;\n          console.log(`⚠️ Batch ${i + 1} failed (attempt ${retryCount}/${maxRetries + 1}): ${batchError.message}`);\n          \n          if (retryCount <= maxRetries) {\n            // Exponential backoff: wait longer between retries\n            const delay = Math.pow(2, retryCount) * 500; // 1s, 2s delays\n            console.log(`⏳ Retrying batch ${i + 1} in ${delay}ms...`);\n            await new Promise(resolve => setTimeout(resolve, delay));\n          } else {\n            console.error(`❌ Batch ${i + 1} failed permanently after ${maxRetries + 1} attempts`);\n            failedBatches++;\n          }\n        }\n      }\n      \n      // Longer delay between batches to avoid rate limiting\n      if (i < batches.length - 1) {\n        await new Promise(resolve => setTimeout(resolve, 300)); // Increased to 300ms\n      }\n    }\n    \n    // Determine success based on how many batches failed\n    const successRate = (batches.length - failedBatches) / batches.length;\n    const isSuccess = successRate >= 0.8; // Consider success if 80%+ of batches succeeded\n    \n    console.log(`📊 Copy operation completed: ${totalCopied}/${blocks.length} blocks copied, ${failedBatches}/${batches.length} batches failed`);\n    \n    return {\n      success: isSuccess,\n      data: { \n        totalBlocks: totalCopied, \n        batches: batches.length,\n        failedBatches: failedBatches,\n        successRate: Math.round(successRate * 100)\n      },\n      timestamp: new Date().toISOString(),\n      warning: failedBatches > 0 ? `${failedBatches} batches failed, but ${Math.round(successRate * 100)}% of content was copied` : undefined\n    };\n  } catch (error) {\n    console.error(`❌ Error copying blocks:`, error);\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\n/**\n * Archive a single page (Notion's way of deleting)\n */\nexport async function archivePage(pageId: string) {\n  try {\n    const response = await notion.pages.update({\n      page_id: pageId,\n      archived: true\n    });\n    \n    return {\n      success: true,\n      data: response,\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\n/**\n * Delete all pages from a database by archiving them\n */\nexport async function deleteAllDatabasePages(databaseId: string) {\n  try {\n    console.log(`🗑️ Getting all pages from database: ${databaseId}`);\n    \n    // Get all pages in the database\n    const pagesResult = await getDatabasePages(databaseId);\n    if (!pagesResult.success) {\n      return {\n        success: false,\n        error: `Failed to get database pages: ${pagesResult.error}`,\n        timestamp: new Date().toISOString(),\n      };\n    }\n    \n    const pages = pagesResult.data.results;\n    console.log(`🗑️ Found ${pages.length} pages to delete`);\n    \n    if (pages.length === 0) {\n      return {\n        success: true,\n        data: { deletedPages: 0 },\n        timestamp: new Date().toISOString(),\n      };\n    }\n    \n    // Archive each page\n    let deletedCount = 0;\n    let failedCount = 0;\n    \n    for (const page of pages) {\n      const archiveResult = await archivePage(page.id);\n      if (archiveResult.success) {\n        deletedCount++;\n        console.log(`✅ Archived page: ${page.id}`);\n      } else {\n        failedCount++;\n        console.error(`❌ Failed to archive page ${page.id}: ${archiveResult.error}`);\n      }\n      \n      // Small delay to avoid rate limiting\n      await new Promise(resolve => setTimeout(resolve, 100));\n    }\n    \n    console.log(`🗑️ Deletion complete: ${deletedCount} deleted, ${failedCount} failed`);\n    \n    return {\n      success: failedCount === 0,\n      data: { \n        deletedPages: deletedCount,\n        failedPages: failedCount,\n        totalPages: pages.length\n      },\n      timestamp: new Date().toISOString(),\n      warning: failedCount > 0 ? `${failedCount} pages failed to delete` : undefined\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\n/**\n * Step 1: Create File Upload Object\n * Creates an upload object in Notion and returns id and upload_url\n */\nexport async function createNotionFileUpload() {\n  try {\n    const apiKey = Deno.env.get(\"NOTION_API_KEY\");\n    \n    const createUploadResponse = await fetch('https://api.notion.com/v1/file_uploads', {\n      method: 'POST',\n      headers: {\n        'Authorization': `Bearer ${apiKey}`,\n        'Content-Type': 'application/json',\n        'Notion-Version': '2022-06-28'\n      },\n      body: JSON.stringify({})\n    });\n\n    if (!createUploadResponse.ok) {\n      const errorText = await createUploadResponse.text();\n      throw new Error(`HTTP ${createUploadResponse.status}: ${errorText}`);\n    }\n\n    const uploadObject = await createUploadResponse.json();\n    \n    return {\n      success: true,\n      data: uploadObject,\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\n/**\n * Step 2: Upload File Contents\n * Uploads the actual file content to Notion's upload URL using FormData\n */\nexport async function uploadFileToNotionUrl(uploadUrl: string, pdfFile: File) {\n  try {\n    const apiKey = Deno.env.get(\"NOTION_API_KEY\");\n    \n    const formData = new FormData();\n    formData.append('file', pdfFile);\n    \n    const uploadResponse = await fetch(uploadUrl, {\n      method: 'POST',\n      headers: {\n        'Authorization': `Bearer ${apiKey}`,\n        'Notion-Version': '2022-06-28'\n        // DO NOT set Content-Type - let FormData handle it\n      },\n      body: formData\n    });\n\n    if (!uploadResponse.ok) {\n      const errorText = await uploadResponse.text();\n      throw new Error(`HTTP ${uploadResponse.status}: ${errorText}`);\n    }\n\n    const uploadResult = await uploadResponse.json();\n    \n    return {\n      success: true,\n      data: uploadResult,\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\n/**\n * Step 3: Attach File to Page\n * Attaches the uploaded file to a Notion page property using file_upload type\n */\nexport async function attachFileUploadToPage(pageId: string, propertyName: string, uploadObjectId: string, fileName: string) {\n  try {\n    const apiKey = Deno.env.get(\"NOTION_API_KEY\");\n    \n    const attachResponse = await fetch(`https://api.notion.com/v1/pages/${pageId}`, {\n      method: 'PATCH',\n      headers: {\n        'Authorization': `Bearer ${apiKey}`,\n        'Notion-Version': '2022-06-28',\n        'Content-Type': 'application/json'\n      },\n      body: JSON.stringify({\n        properties: {\n          [propertyName]: {\n            files: [{\n              name: fileName,\n              type: \"file_upload\",  // NOT \"file\" or \"external\"\n              file_upload: {\n                id: uploadObjectId  // ID from Step 1\n              }\n            }]\n          }\n        }\n      })\n    });\n\n    if (!attachResponse.ok) {\n      const errorText = await attachResponse.text();\n      throw new Error(`HTTP ${attachResponse.status}: ${errorText}`);\n    }\n\n    const attachResult = await attachResponse.json();\n    \n    return {\n      success: true,\n      data: attachResult,\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}",[],[235,448,470],{"_102":477,"_95":478,"_155":479,"_157":449,"_158":137,"_159":480,"_161":162,"_104":481,"_99":160,"_164":260,"_166":482,"_167":22,"_168":-5,"_169":483,"_98":484},"48eaad56-7a0c-11f0-b82f-0224a6c84d84","openai.service.ts","82c4626c-7a05-11f0-9236-0224a6c84d84",23,["D",1755285362994],"/**\n * Two-Stage Consensus Analysis for Meeting Transcripts\n * \n * Stage 1: Analyze each transcript individually against the question\n * Stage 2: Synthesize consensus across all individual analyses\n * \n * Uses actual OpenAI Chat Completions API with proper error handling\n */\n\nimport { OpenAI } from \"https://esm.town/v/std/openai\";\n\n/*** Types ***/\nexport interface ConsensusAnalysisRequest {\n  question: string;\n  transcripts: string[];\n  apiKey: string;\n}\n\nexport type Stance = \"supports\" | \"mixed\" | \"contradicts\" | \"not enough evidence\";\n\nexport interface PerTranscriptFinding {\n  key: string;\n  stance: Stance;\n  shortRationale: string; // <= 120 chars\n}\n\nexport interface ConsensusAnswer {\n  question: string;\n  consensusLevel: \"high\" | \"medium\" | \"low\";\n  directAnswer: string;\n  keyPoints: string[];\n  disagreements: string[];\n  confidence: number; // 0..1\n  methodologyNote: string;\n  perTranscriptFindings: PerTranscriptFinding[];\n}\n\nexport interface ConsensusAnalysisResponse {\n  success: boolean;\n  data?: ConsensusAnswer;\n  error?: string;\n  timestamp: string;\n}\n\nexport interface ConsensusOptions {\n  extractModel?: string; // default: \"gpt-4o-mini\"\n  consensusModel?: string; // default: \"gpt-4o\"\n  maxConcurrency?: number; // default: 3\n  maxTranscriptChars?: number; // default: 15000\n}\n\ninterface IndividualAnalysis {\n  transcriptKey: string;\n  stance: Stance;\n  keyFindings: string[];\n  supportingQuotes: string[];\n  rationale: string;\n  confidence: number;\n}\n\n/*** Configuration ***/\nconst DEFAULTS: Required = {\n  extractModel: \"gpt-4o-mini\",\n  consensusModel: \"gpt-4o\",\n  maxConcurrency: 3,\n  maxTranscriptChars: 100000,\n};\n\n/*** Stage 1: Individual Transcript Analysis ***/\nconst INDIVIDUAL_ANALYSIS_PROMPT = `You are a meticulous research analyst. Analyze the provided interview transcript to answer a specific research question.\n\nYour task:\n1. Determine your stance on how this transcript answers the question\n2. Extract key findings that support your stance\n3. Find relevant quotes (if any) that support your findings\n4. Provide a clear rationale for your stance\n5. Rate your confidence in this analysis\n\nResponse format (JSON):\n{\n  \"stance\": \"supports|mixed|contradicts|not enough evidence\",\n  \"keyFindings\": [\"finding 1\", \"finding 2\", \"finding 3\"],\n  \"supportingQuotes\": [\"quote 1\", \"quote 2\"],\n  \"rationale\": \"Brief explanation of why you chose this stance\",\n  \"confidence\": 0.85\n}\n\nGuidelines:\n- Be objective and evidence-based\n- If the transcript doesn't address the question, use \"not enough evidence\"\n- Keep findings concise but specific\n- Quotes should be under 140 characters\n- Confidence should reflect how clearly the transcript addresses the question (0.0 to 1.0)`;\n\nasync function analyzeIndividualTranscript(\n  openai: OpenAI,\n  model: string,\n  question: string,\n  transcript: string,\n  transcriptKey: string,\n  maxChars: number\n): Promise {\n  // Truncate transcript if too long\n  const truncatedTranscript = transcript.length > maxChars \n    ? transcript.substring(0, maxChars) + \"... [truncated]\"\n    : transcript;\n\n  const userPrompt = `RESEARCH QUESTION:\n${question}\n\nTRANSCRIPT (${transcriptKey}):\n${truncatedTranscript}`;\n\n  try {\n    const completion = await openai.chat.completions.create({\n      model,\n      messages: [\n        { role: \"system\", content: INDIVIDUAL_ANALYSIS_PROMPT },\n        { role: \"user\", content: userPrompt }\n      ],\n      response_format: { type: \"json_object\" },\n      temperature: 0.1,\n      max_tokens: 1000\n    });\n\n    const content = completion.choices[0]?.message?.content;\n    if (!content) {\n      throw new Error(\"No response content from OpenAI\");\n    }\n\n    const parsed = JSON.parse(content);\n    \n    return {\n      transcriptKey,\n      stance: parsed.stance || \"not enough evidence\",\n      keyFindings: Array.isArray(parsed.keyFindings) ? parsed.keyFindings : [],\n      supportingQuotes: Array.isArray(parsed.supportingQuotes) ? parsed.supportingQuotes : [],\n      rationale: parsed.rationale || \"No rationale provided\",\n      confidence: typeof parsed.confidence === 'number' ? parsed.confidence : 0.5\n    };\n\n  } catch (error) {\n    console.error(`Error analyzing ${transcriptKey}:`, error);\n    return {\n      transcriptKey,\n      stance: \"not enough evidence\",\n      keyFindings: [],\n      supportingQuotes: [],\n      rationale: `Analysis failed: ${error.message}`,\n      confidence: 0.0\n    };\n  }\n}\n\n/*** Stage 2: Consensus Synthesis ***/\nconst CONSENSUS_SYNTHESIS_PROMPT = `You are a senior research synthesizer. You will receive multiple individual analyses of different transcripts for the same research question. Your job is to synthesize these into a comprehensive consensus answer.\n\nAnalyze the individual findings and create a consensus that:\n1. Provides a direct, actionable answer to the research question\n2. Identifies the level of consensus across transcripts\n3. Highlights key points that most transcripts agree on\n4. Notes any significant disagreements or outliers\n5. Provides an overall confidence score\n6. Explains your methodology\n\nResponse format (JSON):\n{\n  \"consensusLevel\": \"high|medium|low\",\n  \"directAnswer\": \"Clear, direct answer to the research question\",\n  \"keyPoints\": [\"point 1\", \"point 2\", \"point 3\"],\n  \"disagreements\": [\"disagreement 1\", \"disagreement 2\"],\n  \"confidence\": 0.85,\n  \"methodologyNote\": \"Brief explanation of how you reached consensus\"\n}\n\nConsensus Level Guidelines:\n- HIGH: 80%+ of transcripts support the same conclusion\n- MEDIUM: 60-79% agreement, or mixed but trending\n- LOW: <60% agreement, or significant contradictions\n\nConfidence Guidelines:\n- Consider both the consensus level and the quality of evidence\n- Factor in how directly transcripts addressed the question\n- Account for the strength of supporting evidence`;\n\nasync function synthesizeConsensus(\n  openai: OpenAI,\n  model: string,\n  question: string,\n  individualAnalyses: IndividualAnalysis[]\n): Promise {\n  const analysisText = individualAnalyses.map((analysis, index) => \n    `### ${analysis.transcriptKey}\nStance: ${analysis.stance}\nConfidence: ${analysis.confidence}\nKey Findings: ${analysis.keyFindings.join('; ')}\nSupporting Quotes: ${analysis.supportingQuotes.join('; ')}\nRationale: ${analysis.rationale}`\n  ).join('\\n\\n');\n\n  const userPrompt = `RESEARCH QUESTION:\n${question}\n\nINDIVIDUAL TRANSCRIPT ANALYSES:\n${analysisText}\n\nPlease synthesize these analyses into a comprehensive consensus answer.`;\n\n  try {\n    const completion = await openai.chat.completions.create({\n      model,\n      messages: [\n        { role: \"system\", content: CONSENSUS_SYNTHESIS_PROMPT },\n        { role: \"user\", content: userPrompt }\n      ],\n      response_format: { type: \"json_object\" },\n      temperature: 0.1,\n      max_tokens: 1500\n    });\n\n    const content = completion.choices[0]?.message?.content;\n    if (!content) {\n      throw new Error(\"No response content from OpenAI\");\n    }\n\n    const parsed = JSON.parse(content);\n    \n    // Create per-transcript findings for the final output\n    const perTranscriptFindings: PerTranscriptFinding[] = individualAnalyses.map(analysis => ({\n      key: analysis.transcriptKey,\n      stance: analysis.stance,\n      shortRationale: analysis.rationale.length > 120 \n        ? analysis.rationale.substring(0, 117) + \"...\"\n        : analysis.rationale\n    }));\n\n    return {\n      question,\n      consensusLevel: parsed.consensusLevel || \"low\",\n      directAnswer: parsed.directAnswer || \"Unable to determine consensus answer\",\n      keyPoints: Array.isArray(parsed.keyPoints) ? parsed.keyPoints : [],\n      disagreements: Array.isArray(parsed.disagreements) ? parsed.disagreements : [],\n      confidence: typeof parsed.confidence === 'number' ? parsed.confidence : 0.5,\n      methodologyNote: parsed.methodologyNote || \"Consensus analysis completed\",\n      perTranscriptFindings\n    };\n\n  } catch (error) {\n    console.error(\"Error in consensus synthesis:\", error);\n    \n    // Fallback consensus based on individual analyses\n    const stanceCounts = individualAnalyses.reduce((acc, analysis) => {\n      acc[analysis.stance] = (acc[analysis.stance] || 0) + 1;\n      return acc;\n    }, {} as Record);\n\n    const totalAnalyses = individualAnalyses.length;\n    const maxStance = Object.entries(stanceCounts).reduce((a, b) => \n      stanceCounts[a[0] as Stance] > stanceCounts[b[0] as Stance] ? a : b\n    )[0] as Stance;\n    \n    const maxCount = stanceCounts[maxStance];\n    const consensusPercentage = maxCount / totalAnalyses;\n    \n    const consensusLevel = consensusPercentage >= 0.8 ? \"high\" : \n                          consensusPercentage >= 0.6 ? \"medium\" : \"low\";\n\n    const perTranscriptFindings: PerTranscriptFinding[] = individualAnalyses.map(analysis => ({\n      key: analysis.transcriptKey,\n      stance: analysis.stance,\n      shortRationale: analysis.rationale.length > 120 \n        ? analysis.rationale.substring(0, 117) + \"...\"\n        : analysis.rationale\n    }));\n\n    return {\n      question,\n      consensusLevel,\n      directAnswer: `Synthesis failed, but ${maxCount}/${totalAnalyses} transcripts ${maxStance} the question. Error: ${error.message}`,\n      keyPoints: [],\n      disagreements: [`Synthesis error: ${error.message}`],\n      confidence: 0.3,\n      methodologyNote: `Fallback analysis due to synthesis error. Based on stance distribution: ${JSON.stringify(stanceCounts)}`,\n      perTranscriptFindings\n    };\n  }\n}\n\n/*** Concurrency Control ***/\nasync function mapWithLimit(\n  items: T[],\n  limit: number,\n  fn: (item: T, index: number) => Promise\n): Promise {\n  const results: R[] = new Array(items.length);\n  let index = 0;\n\n  const workers = Array.from({ length: Math.min(limit, items.length) }, async () => {\n    while (index < items.length) {\n      const currentIndex = index++;\n      results[currentIndex] = await fn(items[currentIndex], currentIndex);\n    }\n  });\n\n  await Promise.all(workers);\n  return results;\n}\n\n/*** Public API ***/\nexport async function analyzeTranscriptsWithConsensus(\n  request: ConsensusAnalysisRequest,\n  options: ConsensusOptions = {}\n): Promise {\n  const timestamp = new Date().toISOString();\n\n  // Validation\n  if (!request?.apiKey) {\n    return { success: false, error: \"Missing OpenAI API key\", timestamp };\n  }\n  if (!request?.question?.trim()) {\n    return { success: false, error: \"Missing or empty question\", timestamp };\n  }\n  if (!request?.transcripts?.length) {\n    return { success: false, error: \"No transcripts provided\", timestamp };\n  }\n\n  const config = { ...DEFAULTS, ...options };\n  const openai = new OpenAI({ apiKey: request.apiKey });\n\n  try {\n    console.log(`🔍 Starting 2-stage analysis for question: \"${request.question}\"`);\n    console.log(`📄 Processing ${request.transcripts.length} transcripts with max ${config.maxConcurrency} concurrent requests`);\n\n    // Stage 1: Analyze each transcript individually\n    console.log(\"📊 Stage 1: Individual transcript analysis...\");\n    const individualAnalyses = await mapWithLimit(\n      request.transcripts,\n      config.maxConcurrency,\n      async (transcript, index) => {\n        const transcriptKey = `Transcript ${index + 1}`;\n        console.log(`  🔍 Analyzing ${transcriptKey}...`);\n        \n        const analysis = await analyzeIndividualTranscript(\n          openai,\n          config.extractModel,\n          request.question,\n          transcript,\n          transcriptKey,\n          config.maxTranscriptChars\n        );\n        \n        console.log(`  ✅ ${transcriptKey}: ${analysis.stance} (confidence: ${Math.round(analysis.confidence * 100)}%)`);\n        return analysis;\n      }\n    );\n\n    // Stage 2: Synthesize consensus\n    console.log(\"🤝 Stage 2: Consensus synthesis...\");\n    const consensus = await synthesizeConsensus(\n      openai,\n      config.consensusModel,\n      request.question,\n      individualAnalyses\n    );\n\n    console.log(`✅ Consensus reached: ${consensus.consensusLevel} (confidence: ${Math.round(consensus.confidence * 100)}%)`);\n\n    return {\n      success: true,\n      data: consensus,\n      timestamp\n    };\n\n  } catch (error) {\n    console.error(\"❌ Error in consensus analysis:\", error);\n    return {\n      success: false,\n      error: error instanceof Error ? error.message : String(error),\n      timestamp\n    };\n  }\n}\n\n// Legacy function for backward compatibility (not used in new approach)\nexport async function analyzeTranscriptsForQuestion(\n  question: string,\n  transcripts: string[],\n  apiKey: string\n): Promise<{ success: boolean; data?: any; error?: string }> {\n  const result = await analyzeTranscriptsWithConsensus({\n    question,\n    transcripts,\n    apiKey\n  });\n  \n  return {\n    success: result.success,\n    data: result.data ? { answer: result.data.directAnswer } : undefined,\n    error: result.error\n  };\n}",[],[235,448,478],[235,448],{"_102":487,"_95":488,"_155":489,"_157":236,"_158":137,"_159":160,"_161":162,"_104":490,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_169":491,"_98":524},"d0f9df7a-7951-11f0-8810-0224a6c84d84","types","0198aa63-0ee2-725e-8c9f-e81d8e4fe527",["D",1755205275298],[492,500,508,516],{"_102":493,"_95":247,"_155":494,"_157":489,"_158":137,"_159":218,"_161":162,"_104":495,"_99":496,"_164":177,"_166":497,"_167":22,"_168":-5,"_169":498,"_98":499},"fdcab982-841b-11f0-bd95-0224a6c84d84","0198aa63-0ee2-725e-8c9f-e4354f333684",["D",1756391620516],3,"# Types\n\nShared TypeScript type definitions.\n\n## Available Types\n\n- `email.types.ts` - Email processing and attachment handling types\n  - **Email Interface**: Val Town's standard Email interface with File[] attachments\n  - **Processing Types**: Request/response interfaces for email attachment workflows\n  - **Result Types**: Detailed processing results and error tracking\n- `findings.types.ts` - Types for findings analysis workflow\n- `notion.types.ts` - Notion API property type definitions",[],[235,488,247],{"_102":501,"_95":502,"_155":503,"_157":489,"_158":137,"_159":218,"_161":162,"_104":504,"_99":32,"_164":260,"_166":505,"_167":22,"_168":-5,"_169":506,"_98":507},"fdca959c-841b-11f0-bd95-0224a6c84d84","email.types.ts","fdca961e-841b-11f0-bd95-0224a6c84d84",["D",1756391620516],"/**\n * Email Processing Type Definitions\n * TypeScript interfaces for email attachment processing\n */\n\n// Val Town's actual Email interface\nexport interface Email {\n  from: string;\n  to: string[];\n  cc?: string | string[];\n  bcc?: string | string[];\n  subject?: string;\n  text?: string;\n  html?: string;\n  attachments?: File[];  // File[] array, NOT custom objects\n}\n\nexport interface ProcessEmailAttachmentsRequest {\n  attachments: File[];  // Use actual File objects\n  databaseId: string;\n  pdfPropertyName?: string;\n}\n\nexport interface AttachmentProcessingResult {\n  filename: string;\n  pageCreated: boolean;\n  pageId?: string;\n  fileUploaded: boolean;\n  uploadId?: string;\n  error?: string;\n}\n\nexport interface ProcessEmailAttachmentsResponse {\n  success: boolean;\n  totalAttachments: number;\n  processedAttachments: number;\n  successfulUploads: number;\n  results: AttachmentProcessingResult[];\n  errors: string[];\n  timestamp: string;\n}\n\nexport interface NotionFileUploadObject {\n  id: string;\n  upload_url: string;\n  expiry_time: string;\n}",[],[235,488,502],{"_102":509,"_95":510,"_155":511,"_157":489,"_158":137,"_159":268,"_161":162,"_104":512,"_99":379,"_164":260,"_166":513,"_167":22,"_168":-5,"_169":514,"_98":515},"8ab526ae-840b-11f0-a09a-0224a6c84d84","findings.types.ts","82c46dca-7a05-11f0-9236-0224a6c84d84",["D",1756384555490],"/**\n * Types for findings analysis workflow\n */\n\nexport interface FindingsQuestion {\n  id: string;\n  name: string; // The question text from Name property\n}\n\nexport interface TranscriptPage {\n  id: string;\n  blobKey: string; // From Blob key property\n}\n\nexport interface TranscriptBlob {\n  pageId: string;\n  extractedText: string;\n  extractedAt: string;\n  textLength: number;\n  pdfUrl: string;\n}\n\nexport interface QuestionAnalysis {\n  question: string;\n  answer: string;\n}\n\n// Enhanced consensus analysis types\nexport interface ConsensusAnswer {\n  question: string;\n  consensusLevel: \"high\" | \"medium\" | \"low\";\n  directAnswer: string;\n  keyPoints: string[];              // most agreed-on points\n  disagreements: string[];          // notable divergences/outliers\n  confidence: number;               // 0-1\n  methodologyNote: string;          // how consensus was computed\n  perTranscriptFindings: Array<{\n    key: string;                    // transcript identifier\n    stance: \"supports\" | \"mixed\" | \"contradicts\" | \"not enough evidence\";\n    shortRationale: string;         // one-liner why\n  }>;\n}\n\nexport interface EnhancedQuestionAnalysis {\n  question: string;\n  consensusAnswer: ConsensusAnswer;\n}\n\nexport interface ProcessFindingsRequest {\n  // No parameters needed - uses environment variables\n}\n\nexport interface ProcessFindingsResponse {\n  success: boolean;\n  reportPageId?: string;\n  questionsProcessed?: number;\n  questionsSucceeded?: number;\n  transcriptsAnalyzed?: number;\n  remainingQuestions?: number; // Number of questions still to be processed\n  timestamp: string;\n  error?: string;\n  details?: string;\n}\n\nexport interface DuplicatePageRequest {\n  pageId: string;\n}\n\nexport interface DuplicatePageResponse {\n  success: boolean;\n  sourcePageId: string;\n  targetPageId?: string;\n  blocksCopied?: number;\n  blocksFiltered?: number;\n  timestamp: string;\n  error?: string;\n  details?: string;\n  warning?: string;\n}\n\nexport interface ConvertPageRequest {\n  pageId: string;\n}\n\nexport interface ConvertPageResponse {\n  success: boolean;\n  sourcePageId: string;\n  targetDatabaseId: string;\n  h2SectionsFound: number;\n  pagesCreated: number;\n  pagesWithContent: number;\n  timestamp: string;\n  error?: string;\n  details?: string;\n  warnings?: string[];\n}\n\nexport interface H2Section {\n  title: string;\n  contentBlocks: any[]; // excludes the H2 block itself\n}\n\nexport interface DuplicatePageWithFilteredDatabaseRequest {\n  pageId: string;\n}\n\nexport interface DuplicatePageWithFilteredDatabaseResponse {\n  success: boolean;\n  sourcePageId: string;\n  targetPageId?: string;\n  childDatabaseId?: string;\n  targetChildDatabaseId?: string;\n  childPagesProcessed?: number;\n  childPagesSucceeded?: number;\n  timestamp: string;\n  error?: string;\n  warnings?: string[];\n}",[],[235,488,510],{"_102":517,"_95":518,"_155":519,"_157":489,"_158":137,"_159":218,"_161":162,"_104":520,"_99":160,"_164":260,"_166":521,"_167":22,"_168":-5,"_169":522,"_98":523},"fdcabb12-841b-11f0-bd95-0224a6c84d84","notion.types.ts","8ab4c2a4-840b-11f0-a09a-0224a6c84d84",["D",1756391620516],"/**\n * Notion API Type Definitions\n * TypeScript interfaces for Notion property structures\n */\n\nexport interface NotionDateProperty {\n  date: {\n    start: string;\n  };\n}\n\nexport interface NotionSelectProperty {\n  select: {\n    name: string;\n  };\n}\n\nexport interface NotionRichTextProperty {\n  rich_text: Array<{\n    text: {\n      content: string;\n    };\n  }>;\n}\n\nexport interface NotionNumberProperty {\n  number: number;\n}\n\nexport interface NotionFileProperty {\n  files: Array<{\n    name: string;\n    type: \"file_upload\";\n    file_upload: {\n      id: string;\n    };\n  }>;\n}\n\nexport interface NotionProperties {\n  [key: string]: NotionDateProperty | NotionSelectProperty | NotionRichTextProperty | NotionNumberProperty | NotionFileProperty;\n}",[],[235,488,518],[235,488],{"_102":526,"_95":527,"_155":528,"_157":236,"_158":137,"_159":160,"_161":162,"_104":529,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_169":530,"_98":587},"d0f9e376-7951-11f0-8810-0224a6c84d84","utils","0198aa63-0ee2-725e-8c9f-f5ea225718eb",["D",1755205275298],[531,538,546,554,562,570,579],{"_102":532,"_95":247,"_155":533,"_157":528,"_158":137,"_159":249,"_161":162,"_104":534,"_99":287,"_164":177,"_166":535,"_167":22,"_168":-5,"_169":536,"_98":537},"72e03fbc-8426-11f0-ad01-0224a6c84d84","0198aa63-0ee2-725e-8c9f-ee81b2256350",["D",1756396111921],"# Utils\n\nCommon utility functions.\n\n## Available Utils\n\n- `email.utils.ts` - Email processing utilities (filename parsing, interviewee name extraction)\n- `pdfExtractor.ts` - PDF text extraction utilities\n- `process-timing.utils.ts` - Process timing calculations and formatting utilities\n- `notion-properties.utils.ts` - Notion property creation and formatting utilities\n  - `createTotalQuestionsProperty(count)` - Creates number property for total questions count\n  - `createCompletedQuestionsProperty(count)` - Creates number property for completed questions count\n  - `extractPageIdFromWebhook(webhookBody)` - Extracts page ID from Notion webhook payload\n  - `extractFullNameFromPage(page)` - Extracts full name from Notion page Name property\n- `text-redaction.utils.ts` - Text redaction utilities for sensitive information\n  - `redactPersonName(text, fullName)` - Redacts person's name from text, replacing with \"Respondent\"",[],[235,527,247],{"_102":539,"_95":540,"_155":541,"_157":528,"_158":137,"_159":218,"_161":162,"_104":542,"_99":32,"_164":260,"_166":543,"_167":22,"_168":-5,"_169":544,"_98":545},"fdca9948-841b-11f0-bd95-0224a6c84d84","email.utils.ts","fdca99c0-841b-11f0-bd95-0224a6c84d84",["D",1756391620516],"/**\n * Email Utility Functions\n * Utility functions for processing email data and attachments\n */\n\n/**\n * Extract interviewee name from filename\n * Looks for content in parentheses, falls back to filename without extension\n * \n * @param filename - The original filename from email attachment\n * @returns Extracted interviewee name or filename without extension\n * \n * @example\n * extractIntervieweeName(\"Dealfront discussion (Anna Moch) - 2025_08_04.pdf\")\n * // Returns: \"Anna Moch\"\n * \n * extractIntervieweeName(\"meeting-notes.pdf\")\n * // Returns: \"meeting-notes\"\n */\nexport function extractIntervieweeName(filename: string): string {\n  // Look for content in parentheses\n  const match = filename.match(/\\(([^)]+)\\)/);\n  if (match) {\n    return match[1].trim(); // Return content inside first parentheses\n  }\n  \n  // Fallback: use filename without extension\n  return filename.replace(/\\.[^/.]+$/, \"\").trim();\n}",[],[235,527,540],{"_102":547,"_95":548,"_155":549,"_157":528,"_158":137,"_159":249,"_161":162,"_104":550,"_99":287,"_164":260,"_166":551,"_167":22,"_168":-5,"_169":552,"_98":553},"72e041a6-8426-11f0-ad01-0224a6c84d84","notion-properties.utils.ts","82c47040-7a05-11f0-9236-0224a6c84d84",["D",1756396111921],"/**\n * Notion Properties Utilities\n * Helper functions for creating Notion property objects\n */\n\nimport type {\n  NotionDateProperty,\n  NotionSelectProperty,\n  NotionRichTextProperty,\n  NotionNumberProperty,\n  NotionProperties,\n} from \"../types/notion.types.ts\";\n\n/**\n * Create a Notion date property\n */\nexport function createDateProperty(date: Date): NotionDateProperty {\n  return {\n    date: {\n      start: date.toISOString()\n    }\n  };\n}\n\n/**\n * Create a Notion select property\n */\nexport function createStatusProperty(status: string): NotionSelectProperty {\n  return {\n    select: {\n      name: status\n    }\n  };\n}\n\n/**\n * Create a Notion number property for total questions count\n */\nexport function createTotalQuestionsProperty(count: number): NotionNumberProperty {\n  return {\n    number: count\n  };\n}\n\n/**\n * Create a Notion number property for completed questions count\n */\nexport function createCompletedQuestionsProperty(count: number): NotionNumberProperty {\n  return {\n    number: count\n  };\n}\n\n/**\n * Create timestamp properties for process tracking\n */\nexport function createTimestampProperties(\n  start?: Date,\n  end?: Date,\n  status?: string\n): NotionProperties {\n  const properties: NotionProperties = {};\n\n  if (start) {\n    properties[\"Start time\"] = createDateProperty(start);\n  }\n\n  if (end) {\n    properties[\"End time\"] = createDateProperty(end);\n  }\n\n  if (status) {\n    properties[\"Status\"] = createStatusProperty(status);\n  }\n\n  return properties;\n}\n\n/**\n * Create Analysis Results database schema\n * Defines the structure and options for the findings analysis database\n */\nexport function createAnalysisResultsDatabaseSchema(): any {\n  return {\n    \"Name\": {\n      title: {}\n    },\n    \"Consensus Level\": {\n      select: {\n        options: [\n          { name: \"High\", color: \"green\" },\n          { name: \"Medium\", color: \"yellow\" },\n          { name: \"Low\", color: \"red\" }\n        ]\n      }\n    },\n    \"Confidence Score\": {\n      number: {\n        format: \"percent\"\n      }\n    },\n    \"Status\": {\n      select: {\n        options: [\n          { name: \"Done\", color: \"green\" },\n          { name: \"Error\", color: \"red\" }\n        ]\n      }\n    },\n    \"Category\": {\n      select: {\n        options: [\n          { name: \"Strategy\", color: \"blue\" },\n          { name: \"GTM\", color: \"purple\" },\n          { name: \"AI\", color: \"orange\" },\n          { name: \"Product feedback\", color: \"pink\" },\n          { name: \"Competition\", color: \"brown\" }\n        ]\n      }\n    },\n    \"Priority\": {\n      select: {\n        options: [\n          { name: \"1\", color: \"red\" },\n          { name: \"2\", color: \"yellow\" },\n          { name: \"3\", color: \"green\" },\n          { name: \"Exclude\", color: \"gray\" }\n        ]\n      }\n    },\n    \"Evidence Summary\": {\n      rich_text: {}\n    },\n    \"Methodology\": {\n      rich_text: {}\n    }\n  };\n}\n\n/**\n * Create Results Database ID property for storing database reference\n */\nexport function createResultsDatabaseProperty(databaseId: string): NotionProperties {\n  return {\n    \"Results Database ID\": {\n      rich_text: [\n        {\n          text: {\n            content: databaseId\n          }\n        }\n      ]\n    } as NotionRichTextProperty\n  };\n}\n\n/**\n * Create Evidence Summary property for storing evidence table\n */\nexport function createEvidenceSummaryProperty(evidenceText: string): NotionRichTextProperty {\n  return {\n    rich_text: [\n      {\n        text: {\n          content: evidenceText.substring(0, 2000) // Notion rich text limit\n        }\n      }\n    ]\n  };\n}\n\n/**\n * Create Methodology property for storing methodology note\n */\nexport function createMethodologyProperty(methodologyText: string): NotionRichTextProperty {\n  return {\n    rich_text: [\n      {\n        text: {\n          content: methodologyText.substring(0, 2000) // Notion rich text limit\n        }\n      }\n    ]\n  };\n}\n\n/**\n * Extract page ID from Notion webhook payload\n */\nexport function extractPageIdFromWebhook(webhookBody: any): string | null {\n  const pageId = webhookBody.data?.id;\n  if (!pageId) {\n    console.error(\n      \"❌ No page ID found in webhook payload\",\n      JSON.stringify(webhookBody, null, 2)\n    );\n    return null;\n  }\n  return pageId;\n}\n\n/**\n * Extract full name from Notion page Name property\n */\nexport function extractFullNameFromPage(page: any): string | null {\n  const nameProperty = page.properties?.Name;\n  if (nameProperty?.type === 'title' && nameProperty.title?.length > 0) {\n    return nameProperty.title\n      .map((text: any) => text.plain_text || text.text?.content || \"\")\n      .join(\"\")\n      .trim();\n  }\n  return null;\n}",[],[235,527,548],{"_102":555,"_95":556,"_155":557,"_157":528,"_158":137,"_159":227,"_161":162,"_104":558,"_99":32,"_164":260,"_166":559,"_167":22,"_168":-5,"_169":560,"_98":561},"c8fa4844-79c6-11f0-95e0-0224a6c84d84","pdfExtractor.ts","c8fa48c6-79c6-11f0-95e0-0224a6c84d84",["D",1755255513063],"/**\n * PDF Text Extraction Utility\n * Extracts text from PDF files using pdfjs-dist\n */\n\nexport async function extractTextFromPDF(file: File): Promise {\n  try {\n    console.log(`📄 Starting PDF text extraction for file: ${file.name}`);\n    \n    // Convert File to ArrayBuffer\n    const arrayBuffer = await file.arrayBuffer();\n    \n    // Try using pdfjs-dist which is more Deno-compatible\n    const pdfjsLib = await import(\"https://esm.sh/pdfjs-dist@4.0.379/legacy/build/pdf.mjs\");\n    \n    // Load the PDF document\n    const loadingTask = pdfjsLib.getDocument({ data: arrayBuffer });\n    const pdfDocument = await loadingTask.promise;\n    \n    console.log(`📄 PDF loaded with ${pdfDocument.numPages} pages`);\n    \n    // Extract text from all pages\n    let fullText = '';\n    for (let pageNum = 1; pageNum <= pdfDocument.numPages; pageNum++) {\n      const page = await pdfDocument.getPage(pageNum);\n      const textContent = await page.getTextContent();\n      \n      // Combine text items into a single string\n      const pageText = textContent.items\n        .map((item: any) => item.str)\n        .join(' ');\n      \n      fullText += pageText + '\\n';\n    }\n    \n    console.log(`✅ Text extracted successfully. Length: ${fullText.length} characters`);\n    console.log(`📝 Text preview (first 200 chars): ${fullText.substring(0, 200)}...`);\n    \n    return fullText.trim();\n    \n  } catch (error) {\n    console.error(`❌ PDF text extraction failed:`, error);\n    throw new Error(`Failed to extract text from PDF: ${error.message}`);\n  }\n}\n\nexport async function extractTextFromPDFUrl(pdfUrl: string): Promise {\n  try {\n    console.log(`📄 Downloading PDF from URL: ${pdfUrl}`);\n    \n    // Download the PDF file\n    const response = await fetch(pdfUrl);\n    if (!response.ok) {\n      throw new Error(`Failed to download PDF: ${response.status} ${response.statusText}`);\n    }\n    \n    // Convert response to File object\n    const arrayBuffer = await response.arrayBuffer();\n    const file = new File([arrayBuffer], 'document.pdf', { type: 'application/pdf' });\n    \n    // Extract text using the existing function\n    return await extractTextFromPDF(file);\n    \n  } catch (error) {\n    console.error(`❌ PDF download and extraction failed:`, error);\n    throw new Error(`Failed to download and extract text from PDF: ${error.message}`);\n  }\n}",[],[235,527,556],{"_102":563,"_95":564,"_155":565,"_157":528,"_158":137,"_159":209,"_161":162,"_104":566,"_99":32,"_164":260,"_166":567,"_167":22,"_168":-5,"_169":568,"_98":569},"82c47144-7a05-11f0-9236-0224a6c84d84","process-timing.utils.ts","82c471a8-7a05-11f0-9236-0224a6c84d84",["D",1755282453562],"/**\n * Process Timing Utilities\n * Common functions for timing calculations and formatting\n */\n\nexport interface Duration {\n  minutes: number;\n  seconds: number;\n}\n\n/**\n * Format date in NYC timezone\n */\nexport function formatNYCTime(date: Date, format: '12h' | '24h' = '12h'): string {\n  return date.toLocaleString(\"en-US\", {\n    timeZone: \"America/New_York\",\n    year: \"numeric\",\n    month: \"2-digit\",\n    day: \"2-digit\",\n    hour: \"2-digit\",\n    minute: \"2-digit\",\n    second: \"2-digit\",\n    hour12: format === '12h'\n  });\n}\n\n/**\n * Calculate duration between two dates\n */\nexport function calculateDuration(startTime: Date, endTime: Date): Duration {\n  const totalTimeMs = endTime.getTime() - startTime.getTime();\n  const minutes = Math.floor(totalTimeMs / 60000);\n  const seconds = Math.floor((totalTimeMs % 60000) / 1000);\n  \n  return { minutes, seconds };\n}\n\n/**\n * Format duration for display\n */\nexport function formatDuration(duration: Duration): string {\n  return duration.minutes > 0 ? `${duration.minutes}m ${duration.seconds}s` : `${duration.seconds}s`;\n}\n\n/**\n * Create timing summary blocks for Notion report\n */\nexport function createTimingSummaryBlocks(\n  duration: Duration, \n  totalQuestions: number, \n  successfulQuestions: number\n): any[] {\n  const timeDisplay = formatDuration(duration);\n  const successRate = Math.round((successfulQuestions / totalQuestions) * 100);\n  const nycTime = formatNYCTime(new Date(), '12h');\n  \n  return [\n    {\n      object: \"block\",\n      type: \"divider\",\n      divider: {}\n    },\n    {\n      object: \"block\",\n      type: \"heading_3\",\n      heading_3: {\n        rich_text: [\n          {\n            type: \"text\",\n            text: {\n              content: \"📊 Report Generation Summary\"\n            }\n          }\n        ]\n      }\n    },\n    {\n      object: \"block\",\n      type: \"callout\",\n      callout: {\n        icon: { emoji: \"⏱️\" },\n        rich_text: [\n          {\n            type: \"text\",\n            text: {\n              content: `Total time to generate report: ${timeDisplay}`\n            },\n            annotations: { bold: true }\n          }\n        ]\n      }\n    },\n    {\n      object: \"block\",\n      type: \"paragraph\",\n      paragraph: {\n        rich_text: [\n          {\n            type: \"text\",\n            text: {\n              content: `Questions processed: ${successfulQuestions}/${totalQuestions} (${successRate}% success rate)`\n            }\n          }\n        ]\n      }\n    },\n    {\n      object: \"block\",\n      type: \"paragraph\",\n      paragraph: {\n        rich_text: [\n          {\n            type: \"text\",\n            text: {\n              content: `Generated on: ${nycTime} (NYC)`\n            },\n            annotations: { italic: true, color: \"gray\" }\n          }\n        ]\n      }\n    }\n  ];\n}",[],[235,527,564],{"_102":571,"_95":572,"_155":573,"_157":528,"_158":137,"_159":574,"_161":162,"_104":575,"_99":303,"_164":260,"_166":576,"_167":22,"_168":-5,"_169":577,"_98":578},"a58db7a8-8450-11f0-8f00-0224a6c84d84","text-redaction.utils.ts","72e011f4-8426-11f0-ad01-0224a6c84d84",44,["D",1756414235808],"/**\n * Text Redaction Utilities\n * Functions for redacting sensitive information from text content\n */\n\n/**\n * Generate a simple 4-character hash from a string\n */\nfunction generateSimpleHash(str: string): string {\n  let hash = 0;\n  for (let i = 0; i < str.length; i++) {\n    const char = str.charCodeAt(i);\n    hash = ((hash << 5) - hash) + char;\n    hash = hash & hash; // Convert to 32-bit integer\n  }\n  \n  // Convert to positive number and create 4-character alphanumeric string\n  const positiveHash = Math.abs(hash);\n  const base36 = positiveHash.toString(36).toUpperCase();\n  \n  // Pad or truncate to 4 characters\n  return base36.length >= 4 ? base36.substring(0, 4) : base36.padStart(4, '0');\n}\n\n/**\n * Redact person's name from text, replacing with \"Respondent_HASH\"\n * Handles both speaker identification (Name:) and narrative mentions\n */\nexport function redactPersonName(text: string, fullName: string): string {\n  if (!fullName || !text) {\n    return text;\n  }\n\n  // Generate unique hash for this person\n  const hash = generateSimpleHash(fullName);\n  const respondentId = `Respondent_${hash}`;\n\n  // Escape special regex characters in the name\n  const escapedName = fullName.replace(/[.*+?^${}()|[\\]\\\\]/g, '\\\\$&');\n  \n  // Handle \"Name:\" pattern first (speaker identification)\n  const nameWithColonPattern = new RegExp(`\\\\b${escapedName}\\\\s*:`, 'gi');\n  let redactedText = text.replace(nameWithColonPattern, `${respondentId}:`);\n  \n  // Handle name in regular text\n  const nameInTextPattern = new RegExp(`\\\\b${escapedName}\\\\b`, 'gi');\n  redactedText = redactedText.replace(nameInTextPattern, respondentId);\n  \n  return redactedText;\n}",[],[235,527,572],{"_102":580,"_95":581,"_155":582,"_157":528,"_158":137,"_159":160,"_161":162,"_104":583,"_99":32,"_164":260,"_166":584,"_167":22,"_168":-5,"_169":585,"_98":586},"d0f9e240-7951-11f0-8810-0224a6c84d84","url.ts","0198aa63-0ee2-725e-8c9f-f19b9262c162",["D",1755205275298],"export function getAbsoluteUrl(c: any, path: string): string {\n  const protocol = c.req.header(\"x-forwarded-proto\") || \"https\";\n  const host = c.req.header(\"host\");\n  return `${protocol}://${host}${path}`;\n}",[],[235,527,581],[235,527],{"_102":589,"_95":247,"_155":590,"_157":236,"_158":137,"_159":218,"_161":162,"_104":591,"_99":303,"_164":177,"_166":592,"_167":22,"_168":-5,"_169":593,"_98":594},"fdcaa03c-841b-11f0-bd95-0224a6c84d84","0198aa63-0ee2-725e-8c9f-f8f1bff870bc",["D",1756391620516],"# Backend\n\nThis directory contains all server-side code organized by functionality:\n\n- `/routes` - HTTP route handlers organized by functionality\n- `/api` - JSON API endpoints for frontend/backend communication  \n- `/tasks` - Webhook handlers for Notion integrations (PDF extraction, page duplication)\n- `/views` - Server-side rendered views for frontend\n- `/emails` - Email trigger handlers for processing incoming emails\n- `/services` - External API integrations (pure API calls only)\n- `/controllers` - Business logic coordination between routes and services\n- `/crons` - Scheduled task handlers\n- `/types` - Shared TypeScript type definitions\n- `/utils` - Common utility functions\n\n## Architecture\n\n- **Services**: Pure API calls to external systems\n- **Controllers**: Business logic and orchestration\n- **Routes**: HTTP request/response handling only",[],[235,247],[235],{"_102":597,"_95":598,"_155":599,"_157":-5,"_158":137,"_159":160,"_161":162,"_104":600,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_169":601,"_98":609},"d0f9ea2e-7951-11f0-8810-0224a6c84d84","frontend","0198aa63-0ee2-725e-8ca0-054e1ef4f005",["D",1755205275298],[602],{"_102":603,"_95":247,"_155":604,"_157":599,"_158":137,"_159":160,"_161":162,"_104":605,"_99":32,"_164":177,"_166":606,"_167":22,"_168":-5,"_169":607,"_98":608},"d0f9e934-7951-11f0-8810-0224a6c84d84","0198aa63-0ee2-725e-8ca0-01af51fe4bc0",["D",1755205275298],"# Frontend\n\nClient-side files that render in the browser.",[],[598,247],[598],{"_102":611,"_95":612,"_155":613,"_157":-5,"_158":137,"_159":160,"_161":162,"_104":614,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_169":615,"_98":623},"d0f9ec40-7951-11f0-8810-0224a6c84d84","shared","0198aa63-0ee2-725e-8ca0-0c5d990abef8",["D",1755205275298],[616],{"_102":617,"_95":247,"_155":618,"_157":613,"_158":137,"_159":160,"_161":162,"_104":619,"_99":32,"_164":177,"_166":620,"_167":22,"_168":-5,"_169":621,"_98":622},"d0f9eb46-7951-11f0-8810-0224a6c84d84","0198aa63-0ee2-725e-8ca0-0b7410f5f5a9",["D",1755205275298],"# Shared\n\nFiles used by both backend and frontend. Code in this directory must work in both environments.",[],[612,247],[612],{"_102":625,"_95":626,"_155":627,"_157":-5,"_158":137,"_159":160,"_161":162,"_104":628,"_99":32,"_164":177,"_166":629,"_167":22,"_168":-5,"_169":630,"_98":631},"d0f9ed30-7951-11f0-8810-0224a6c84d84",".vtignore","0198aa63-0ee2-725e-8ca0-105d2f2cb6c3",["D",1755205275298],".git\n.vscode\n.cursorrules\n.DS_Store\nnode_modules\nvendor\nvars.env",[],[626],{"_102":633,"_95":634,"_155":635,"_157":-5,"_158":137,"_159":160,"_161":162,"_104":636,"_99":32,"_164":177,"_166":637,"_167":22,"_168":-5,"_169":638,"_98":639},"d0f9ee20-7951-11f0-8810-0224a6c84d84","deno.json","0198aa63-0ee2-725e-8ca0-17f855c1ff92",["D",1755205275298],"{\n  \"$schema\": \"https://raw.githubusercontent.com/denoland/deno/348900b8b79f4a434cab4c74b3bc8d4d2fa8ee74/cli/schemas/config-file.v1.json\",\n  \"lock\": false,\n  \"compilerOptions\": {\n    \"noImplicitAny\": false,\n    \"strict\": false,\n    \"types\": [\n      \"https://www.val.town/types/valtown.d.ts\"\n    ],\n    \"lib\": [\n      \"dom\",\n      \"dom.iterable\",\n      \"dom.asynciterable\",\n      \"deno.ns\",\n      \"deno.unstable\"\n    ]\n  },\n  \"lint\": {\n    \"files\": {\n      \"include\": [\n        \"deno:/https/esm.town/**/*\"\n      ]\n    },\n    \"rules\": {\n      \"exclude\": [\n        \"no-explicit-any\"\n      ]\n    }\n  },\n  \"node_modules_dir\": false,\n  \"experimental\": {\n    \"unstable-node-globals\": true,\n    \"unstable-temporal\": true,\n    \"unstable-worker-options\": true,\n    \"unstable-sloppy-imports\": true\n  }\n}",[],[634],{"_102":641,"_95":642,"_155":643,"_157":-5,"_158":137,"_159":160,"_161":162,"_104":644,"_99":32,"_164":645,"_166":646,"_167":22,"_168":96,"_169":647,"_98":648},"d0f9ef1a-7951-11f0-8810-0224a6c84d84","main.tsx","0198aa63-0ee2-725e-8ca0-1891f44fbfbe",["D",1755205275298],"http","import { Hono } from \"npm:hono@3.12.12\";\nimport { lastlogin } from \"https://esm.town/v/stevekrouse/lastlogin_safe\";\n\n// Import route modules\nimport authCheck from \"./backend/routes/authCheck.ts\";\nimport webhookAuth from \"./backend/routes/webhookAuthCheck.ts\";\nimport dashboardRoute from \"./backend/routes/views/dashboard.tsx\";\nimport apiRoutes from \"./backend/routes/api/_api.routes.ts\";\nimport taskRoutes from \"./backend/routes/tasks/_tasks.routes.ts\";\nimport viewRoutes from \"./backend/routes/views/_views.routes.ts\";\n\nconst app = new Hono();\n\n// Unwrap Hono errors to see original error details\napp.onError((err, c) => {\n  throw err;\n});\n\n// 1. Webhook auth - ONLY for /tasks/* routes\napp.use(\"/tasks/*\", webhookAuth);\n\n// 2. User auth - for ALL routes EXCEPT /tasks/*\napp.use(\"/api/*\", authCheck);\napp.use(\"/views/*\", authCheck);\napp.use(\"/\", authCheck);\n\n// Dashboard route (authenticated)\napp.get(\"/\", dashboardRoute);\n\n// 3. Mount routes AFTER middleware\napp.route(\"/api\", apiRoutes);\napp.route(\"/tasks\", taskRoutes);\napp.route(\"/views\", viewRoutes);\n\n// Main HTTP handler\nasync function handler(request: Request): Promise {\n  return app.fetch(request);\n}\n\n// Wrap and export the main HTTP handler with lastlogin\nexport default lastlogin(handler);",[],[642],"files",[152,171,181,189,197,205,214,223,233,239,245,255,264,273,282,292,298,307,317,323,330,341,347,353,360,369,375,383,392,398,405,413,422,429,437,446,452,459,468,476,486,492,500,508,516,525,531,538,546,554,562,570,579,588,596,602,610,616,624,632,640],"counts",{"_653":32,"_654":303,"_655":32,"_656":209},"pulls","branches","forks","envVars","contents",[659,662,665,668,671,674,677],{"_102":153,"_95":154,"_155":156,"_157":-5,"_158":137,"_159":160,"_161":162,"_104":660,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_98":661},["D",1755205275298],[154],{"_102":234,"_95":235,"_155":236,"_157":-5,"_158":137,"_159":160,"_161":162,"_104":663,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_98":664},["D",1755205275298],[235],{"_102":597,"_95":598,"_155":599,"_157":-5,"_158":137,"_159":160,"_161":162,"_104":666,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_98":667},["D",1755205275298],[598],{"_102":611,"_95":612,"_155":613,"_157":-5,"_158":137,"_159":160,"_161":162,"_104":669,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_98":670},["D",1755205275298],[612],{"_102":625,"_95":626,"_155":627,"_157":-5,"_158":137,"_159":160,"_161":162,"_104":672,"_99":32,"_164":177,"_166":629,"_167":22,"_168":-5,"_98":673},["D",1755205275298],[626],{"_102":633,"_95":634,"_155":635,"_157":-5,"_158":137,"_159":160,"_161":162,"_104":675,"_99":32,"_164":177,"_166":637,"_167":22,"_168":-5,"_98":676},["D",1755205275298],[634],{"_102":641,"_95":642,"_155":643,"_157":-5,"_158":137,"_159":160,"_161":162,"_104":678,"_99":32,"_164":645,"_166":646,"_167":22,"_168":96,"_98":679},["D",1755205275298],[642],"upstreamProject",{"_102":145,"_95":682,"_110":683,"_111":684,"_108":109,"_23":685},"findings-base","Connect Notion and val.town, including user and webhook auth","https://imagedelivery.net/iHX6Ovru0O7AjmyT5yZRoA/2c2c44af-ae1d-44bb-37da-2478b01ac700/square512",{"_93":94,"_129":130},"upstreamBranch",{"_102":140,"_104":688,"_689":690,"_95":43,"_138":145,"_99":142,"_139":-5,"_141":-5,"_691":692,"_106":-5,"_149":22,"_108":109,"_148":94,"_16":118,"_147":682},["D",1755176650069],"updatedAt",["D",1755192248891],"lockfileKey","1755705837344-esm.town_v_lightweight_findings-base@173-main","emptyEnvVars","isNew","flags",[],"routes/_app.x.$handle.$name.compare.$branch",{"_93":94,"_95":96,"_97":43,"_100":699,"_704":334,"_705":32,"_706":707,"_956":162,"_957":703,"_958":959,"_962":-5},{"_102":103,"_95":96,"_104":700,"_106":-5,"_107":22,"_108":109,"_110":-5,"_111":112,"_14":113,"_114":-5,"_115":22,"_116":-5,"_117":32,"_16":118,"_119":120,"_121":120,"_122":123,"_124":123,"_125":123,"_126":120,"_127":120,"_23":701,"_131":702,"_134":703,"_149":22},["D",1755205275298],{"_93":94,"_129":130},{"_93":94,"_133":130},{"_99":136,"_95":43,"_102":137,"_138":103,"_139":140,"_141":142,"_143":109,"_144":145,"_146":118,"_147":96,"_148":94},"ahead","behind","diff",{"_708":709,"_167":812,"_813":814},"added",[710,718,723,728,733,738,743,748,752,757,762,767,772,777,782,787,792,797,802,807],{"_98":711,"_712":-7,"_713":714},"_townie/2-stage-questions-openai.md","targetFile","sourceFile",{"_102":206,"_95":207,"_155":208,"_157":156,"_158":137,"_159":209,"_161":162,"_104":210,"_99":32,"_164":177,"_166":211,"_167":22,"_168":-5,"_169":-7,"_98":715,"_716":717},[154,207],"contentHash","98d74ba81054604bfbe2b59c4e2b0589f1f2ee37193860a3117bdfa7cb4d71de",{"_98":719,"_712":-7,"_713":720},"_townie/email-PDF-attachments.md",{"_102":215,"_95":216,"_155":217,"_157":156,"_158":137,"_159":218,"_161":162,"_104":219,"_99":32,"_164":177,"_166":220,"_167":22,"_168":-5,"_169":-7,"_98":721,"_716":722},[154,216],"919d0ceaf5ef0613244d5fed76c560ed40342a4c04842224641826407f092f5a",{"_98":724,"_712":-7,"_713":725},"_townie/pdf-to-blob.md",{"_102":224,"_95":225,"_155":226,"_157":156,"_158":137,"_159":227,"_161":162,"_104":228,"_99":32,"_164":177,"_166":229,"_167":22,"_168":-5,"_169":-7,"_98":726,"_716":727},[154,225],"8b8b1432cfe39824213f2f243daa837273fa20ef3d982178dc481a10a0e2d7b9",{"_98":729,"_712":-7,"_713":730},"backend/controllers/emails.controller.ts",{"_102":256,"_95":257,"_155":258,"_157":242,"_158":137,"_159":218,"_161":162,"_104":259,"_99":32,"_164":260,"_166":261,"_167":22,"_168":-5,"_169":-7,"_98":731,"_716":732},[235,241,257],"3004fa002b5ecb05b1612800e73bead3510164c5e461c48729f0b124e26ddf03",{"_98":734,"_712":-7,"_713":735},"backend/controllers/findings.controller.ts",{"_102":265,"_95":266,"_155":267,"_157":242,"_158":137,"_159":268,"_161":162,"_104":269,"_99":209,"_164":260,"_166":270,"_167":22,"_168":-5,"_169":-7,"_98":736,"_716":737},[235,241,266],"35da08b7f643302ea304a7c3e8187f87af7454d51360a030a1396aa917a833f9",{"_98":739,"_712":-7,"_713":740},"backend/controllers/tasks.controller.ts",{"_102":283,"_95":284,"_155":285,"_157":242,"_158":137,"_159":249,"_161":162,"_104":286,"_99":287,"_164":260,"_166":288,"_167":22,"_168":-5,"_169":-7,"_98":741,"_716":742},[235,241,284],"57c9faddbbc7b5b2e2637900d4e384edf1c8deb4b0a3dda227da019816cd2f29",{"_98":744,"_712":-7,"_713":745},"backend/crons/findings-analysis.ts",{"_102":308,"_95":309,"_155":310,"_157":295,"_158":137,"_159":301,"_161":162,"_104":311,"_99":160,"_164":312,"_166":313,"_167":22,"_168":-5,"_169":-7,"_98":746,"_716":747},[235,294,309],"93b8bdc25371404186a7de35564be0cd65106658cb277fe41ef9aa5efd111189",{"_98":749,"_712":-7,"_713":750},"backend/emails",{"_102":318,"_95":319,"_155":320,"_157":236,"_158":137,"_159":218,"_161":162,"_104":321,"_99":32,"_164":165,"_166":-5,"_167":22,"_168":-5,"_169":-7,"_98":751,"_716":-5},[235,319],{"_98":753,"_712":-7,"_713":754},"backend/emails/README.md",{"_102":324,"_95":247,"_155":325,"_157":320,"_158":137,"_159":136,"_161":162,"_104":326,"_99":303,"_164":177,"_166":327,"_167":22,"_168":-5,"_169":-7,"_98":755,"_716":756},[235,319,247],"646b214f8c7caa0450d2a6b81244798baecad484aee3767d4ac23c5136d72b2d",{"_98":758,"_712":-7,"_713":759},"backend/emails/pdf-attachments.ts",{"_102":331,"_95":332,"_155":333,"_157":320,"_158":137,"_159":334,"_161":162,"_104":335,"_99":160,"_164":336,"_166":337,"_167":22,"_168":96,"_169":-7,"_98":760,"_716":761},[235,319,332],"16aef39ad80575f2f20f1467811fce33168d335ce5dc8f34ac9d1c2d15a85149",{"_98":763,"_712":-7,"_713":764},"backend/services/email.service.ts",{"_102":460,"_95":461,"_155":462,"_157":449,"_158":137,"_159":463,"_161":162,"_104":464,"_99":303,"_164":260,"_166":465,"_167":22,"_168":-5,"_169":-7,"_98":765,"_716":766},[235,448,461],"085cd4990caa47938946ca28621be88c5a27142a533916d485a7dc485d5e62e9",{"_98":768,"_712":-7,"_713":769},"backend/services/openai.service.ts",{"_102":477,"_95":478,"_155":479,"_157":449,"_158":137,"_159":480,"_161":162,"_104":481,"_99":160,"_164":260,"_166":482,"_167":22,"_168":-5,"_169":-7,"_98":770,"_716":771},[235,448,478],"a1eaa3c32e41a327ba019594e65897296dfea27fe18517c4129cb3dd2ec312e4",{"_98":773,"_712":-7,"_713":774},"backend/types/email.types.ts",{"_102":501,"_95":502,"_155":503,"_157":489,"_158":137,"_159":218,"_161":162,"_104":504,"_99":32,"_164":260,"_166":505,"_167":22,"_168":-5,"_169":-7,"_98":775,"_716":776},[235,488,502],"ccacab015e059973ea3f0af2a1bbe1d8108dd04082d507e2268c0ad069202b2d",{"_98":778,"_712":-7,"_713":779},"backend/types/findings.types.ts",{"_102":509,"_95":510,"_155":511,"_157":489,"_158":137,"_159":268,"_161":162,"_104":512,"_99":379,"_164":260,"_166":513,"_167":22,"_168":-5,"_169":-7,"_98":780,"_716":781},[235,488,510],"0d19005ecaf65dfc087c21957940459dab5f653f02b970f7682a80f6eb3e4982",{"_98":783,"_712":-7,"_713":784},"backend/types/notion.types.ts",{"_102":517,"_95":518,"_155":519,"_157":489,"_158":137,"_159":218,"_161":162,"_104":520,"_99":160,"_164":260,"_166":521,"_167":22,"_168":-5,"_169":-7,"_98":785,"_716":786},[235,488,518],"1aa353afe92b97363c2a5c4603df2673a4e26b56c2b6940ca9ceae51fb4299d9",{"_98":788,"_712":-7,"_713":789},"backend/utils/email.utils.ts",{"_102":539,"_95":540,"_155":541,"_157":528,"_158":137,"_159":218,"_161":162,"_104":542,"_99":32,"_164":260,"_166":543,"_167":22,"_168":-5,"_169":-7,"_98":790,"_716":791},[235,527,540],"c8063d43e7c86cff7bde81533e33ab28d5f7d0283f3216ded17b94a80713115d",{"_98":793,"_712":-7,"_713":794},"backend/utils/notion-properties.utils.ts",{"_102":547,"_95":548,"_155":549,"_157":528,"_158":137,"_159":249,"_161":162,"_104":550,"_99":287,"_164":260,"_166":551,"_167":22,"_168":-5,"_169":-7,"_98":795,"_716":796},[235,527,548],"3114595363a42c99c22c7ecaabc116ac431c42c5abee755ef354da25ac330912",{"_98":798,"_712":-7,"_713":799},"backend/utils/pdfExtractor.ts",{"_102":555,"_95":556,"_155":557,"_157":528,"_158":137,"_159":227,"_161":162,"_104":558,"_99":32,"_164":260,"_166":559,"_167":22,"_168":-5,"_169":-7,"_98":800,"_716":801},[235,527,556],"83c0234c1fbb2fc87e6332ba558a4e7a99076bb2c0c7f9bcb7fda7bd998b5f8d",{"_98":803,"_712":-7,"_713":804},"backend/utils/process-timing.utils.ts",{"_102":563,"_95":564,"_155":565,"_157":528,"_158":137,"_159":209,"_161":162,"_104":566,"_99":32,"_164":260,"_166":567,"_167":22,"_168":-5,"_169":-7,"_98":805,"_716":806},[235,527,564],"8332fe1ccd294ee5a1b68f1ed0aa483b9931ee9844860f1f2230c3b168c3760c",{"_98":808,"_712":-7,"_713":809},"backend/utils/text-redaction.utils.ts",{"_102":571,"_95":572,"_155":573,"_157":528,"_158":137,"_159":574,"_161":162,"_104":575,"_99":303,"_164":260,"_166":576,"_167":22,"_168":-5,"_169":-7,"_98":810,"_716":811},[235,527,572],"bef111218cd0e47009ad6ec930bd77233e4fc235d433d2fc521c6274a493058e",[],"modified",[815,829,842,855,867,880,893,905,918,930,943],{"_98":816,"_712":817,"_713":826},"_townie/00-setup.md",{"_102":818,"_95":173,"_155":819,"_157":820,"_158":140,"_159":821,"_161":162,"_104":822,"_99":287,"_164":177,"_166":823,"_167":22,"_168":-5,"_169":-7,"_98":824,"_716":825},"ee9cde72-792b-11f0-8a95-0224a6c84d84","c6e732ca-7911-11f0-8088-0224a6c84d84","c6c9a688-7911-11f0-8de6-0224a6c84d84",140,["D",1755189004219],"# Setup\n\nUse this file to guide you through the setup process. There isn't much to do but if it's your first time, some concepts may be new, so please follow them carefully and reach out if you have any questions.\n\nThe instructions in this file are for _you_. It's important to distinguish the instructions in this file from the instructions in the other files in this folder, which are for Townie.\n\n## Prompts in this runbook\n\nThe prompts that you will enter into Townie are in blockquotes, like this:\n\n> I'm a prompt for Townie\n\nWhen Townie shows you these, copy and paste them into Townie and run them. Townie will use them to build this Val from the ground up, through these layers and features of the Val:\n\n- scaffolding\n- Notion & val.town integration health endpoint\n- user authentication\n- dashboard\n- webhook authentication\n- caching\n- React front end\n- metrics collection\n\nOnce you're through this runbook, version 1 of this Val will be feature complete and you can add and experiment with new features til your heart's content.\n\n## Assumptions\n\nTo use this runbook, you'll need to have these things ready:\n\n1. A val.town account\n1. A [Remix of this Val](https://docs.val.town/vals/remixes/) in your val.town account\n1. The [environment variables](https://docs.val.town/reference/environment-variables/) imported into your Remix\n\nOnce your environment variables are saved to this Val (i.e., once you've imported `vars.env` to your Remix), all of the environment variables are available to the Val and Townie, and val.town and Notion are connected to each other. (Note that [environment variables are secret and secure on val.town](https://docs.val.town/reference/environment-variables/).)\n\nIf you don't have the environment variables, contact your administrator.\n\n### Environment variables (required)\n\nThis is a look at what the environment variables are, and what they do. (They are not the environment variables themselves; you'll need to get those from your administrator.)\n\nThe most important thing to understand about them is that they are required to connect val.town to Notion, and to connect this Val's code to the Notion databases that hold the demo content.\n\n- `NOTION_API_KEY` - Your Notion integration API key\n- `NOTION_WEBHOOK_SECRET` - Matches the X-API-KEY custom header coming from Notion webhooks to /tasks/\\* endpoint(s)\n- Database environment variables - the keys for the environment variables that hold the ids for the ids for your Notion databases:\n\n  ```yaml\n  project: findings\n  databases:\n  - `FINDINGS_TRANSCRIPTS_DB_ID`\n  - `FINDINGS_QUESTIONS_DB_ID`\n  - `FINDINGS_REPORTS_DB_ID`\n  ```\n\n# Townie instructions\n\nNow that you're through setup, you can instruct Townie to build this Val by pasting prompts into Townie.\n\nIn Townie, select this Val, and prompt Townie with this line:\n\n> Scaffold this Val with what you see in `/_townie/01-scaffold.md`\n\nWhen Townie completes the scaffolding task, it will tell you what you need to do next. Have fun!\n\n## [On to Townie](https://townie.val.run/)!\n\n---\n\n## 02-auth.md\n\n> Add auth to this Val with what you see in `/_townie/02-auth.md`\n\n## 03-webhooks.md\n\n> Add webhook auth to this Val with what you see in `/_townie/03-webhooks.md`\n",[154,173],"2d16a15fee9cdc6d3f194ac430c5d813ffac573d08ec72b870cb5f02ad5650fd",{"_102":172,"_95":173,"_155":174,"_157":156,"_158":137,"_159":175,"_161":162,"_104":176,"_99":160,"_164":177,"_166":178,"_167":22,"_168":-5,"_169":-7,"_98":827,"_716":828},[154,173],"535b631cd9599cc94c80f0e257fdfbe615353999d79337104b8cf1ae536ea4b1",{"_98":830,"_712":831,"_713":839},"backend/README.md",{"_102":832,"_95":247,"_155":833,"_157":834,"_158":140,"_159":142,"_161":162,"_104":835,"_99":32,"_164":177,"_166":836,"_167":22,"_168":-5,"_169":-7,"_98":837,"_716":838},"7c9747ce-7933-11f0-96c0-0224a6c84d84","7c97483c-7933-11f0-96c0-0224a6c84d84","0198a99c-4a38-76d0-9aab-a65fe3077ac9",["D",1755192248891],"# Backend\n\nThis directory contains all server-side code organized by functionality:\n\n- `/routes` - HTTP route handlers organized by functionality\n- `/api` - JSON API endpoints for frontend/backend communication  \n- `/tasks` - Webhook handlers for Notion integrations\n- `/views` - Server-side rendered views for frontend\n- `/services` - External API integrations (pure API calls only)\n- `/controllers` - Business logic coordination between routes and services\n- `/crons` - Scheduled task handlers\n- `/types` - Shared TypeScript type definitions\n- `/utils` - Common utility functions\n\n## Architecture\n\n- **Services**: Pure API calls to external systems\n- **Controllers**: Business logic and orchestration\n- **Routes**: HTTP request/response handling only",[235,247],"19f83049c063057e13f681a192feff321aed57aee886ac5b8c9fa9afe58fbc83",{"_102":589,"_95":247,"_155":590,"_157":236,"_158":137,"_159":218,"_161":162,"_104":591,"_99":303,"_164":177,"_166":592,"_167":22,"_168":-5,"_169":-7,"_98":840,"_716":841},[235,247],"011dfa168341808148d28e5842c706e9845679412e2970cbbb848a51ddb5f3a1",{"_98":843,"_712":844,"_713":852},"backend/controllers/README.md",{"_102":845,"_95":247,"_155":846,"_157":847,"_158":140,"_159":142,"_161":162,"_104":848,"_99":32,"_164":177,"_166":849,"_167":22,"_168":-5,"_169":-7,"_98":850,"_716":851},"7c9743f0-7933-11f0-96c0-0224a6c84d84","7c97447c-7933-11f0-96c0-0224a6c84d84","0198a99c-4a38-76d0-9aab-a8ded408d294",["D",1755192248891],"# Controllers\n\nBusiness logic coordination between routes and services. Controllers should:\n\n- Handle environment variables and configuration\n- Coordinate multiple service calls\n- Implement business rules and validation\n- Handle service business logic layer (result.success)",[235,241,247],"1db9da7de2fb390933bcb1aabc1c97a44837a3fcee45a548798a883626eff4fe",{"_102":246,"_95":247,"_155":248,"_157":242,"_158":137,"_159":249,"_161":162,"_104":250,"_99":251,"_164":177,"_166":252,"_167":22,"_168":-5,"_169":-7,"_98":853,"_716":854},[235,241,247],"ce2ce995c8d2f6a06f2b4314180b34aa1eb558b8e332213e179f0706669177d7",{"_98":856,"_712":857,"_713":864},"backend/controllers/health.controller.ts",{"_102":858,"_95":275,"_155":859,"_157":847,"_158":140,"_159":142,"_161":162,"_104":860,"_99":32,"_164":260,"_166":861,"_167":22,"_168":-5,"_169":-7,"_98":862,"_716":863},"7c973860-7933-11f0-96c0-0224a6c84d84","7c97391e-7933-11f0-96c0-0224a6c84d84",["D",1755192248891],"import { getDatabaseById } from \"../services/notion.service.ts\";\n\nexport async function getConfiguredDatabases() {\n  const transcriptsDbId = Deno.env.get(\"FINDINGS_TRANSCRIPTS_DB_ID\");\n  const questionsDbId = Deno.env.get(\"FINDINGS_QUESTIONS_DB_ID\");\n  const reportsDbId = Deno.env.get(\"FINDINGS_REPORTS_DB_ID\");\n\n  const databases = { \n    transcripts: null, \n    questions: null, \n    reports: null \n  };\n  const errors: string[] = [];\n\n  // Check transcripts database\n  if (transcriptsDbId) {\n    const result = await getDatabaseById(transcriptsDbId);\n    if (result.success) {\n      databases.transcripts = result.data;\n    } else {\n      errors.push(`Transcripts DB: ${result.error}`);\n    }\n  } else {\n    errors.push(\"FINDINGS_TRANSCRIPTS_DB_ID not configured\");\n  }\n\n  // Check questions database\n  if (questionsDbId) {\n    const result = await getDatabaseById(questionsDbId);\n    if (result.success) {\n      databases.questions = result.data;\n    } else {\n      errors.push(`Questions DB: ${result.error}`);\n    }\n  } else {\n    errors.push(\"FINDINGS_QUESTIONS_DB_ID not configured\");\n  }\n\n  // Check reports database\n  if (reportsDbId) {\n    const result = await getDatabaseById(reportsDbId);\n    if (result.success) {\n      databases.reports = result.data;\n    } else {\n      errors.push(`Reports DB: ${result.error}`);\n    }\n  } else {\n    errors.push(\"FINDINGS_REPORTS_DB_ID not configured\");\n  }\n\n  return {\n    success: errors.length === 0,\n    data: databases,\n    errors: errors.length > 0 ? errors : undefined,\n    timestamp: new Date().toISOString(),\n  };\n}\n\nexport async function getHealthStatus() {\n  const configuredDatabases = await getConfiguredDatabases();\n  \n  const healthStatus = {\n    status: configuredDatabases.success ? \"ok\" : \"error\",\n    timestamp: new Date().toISOString(),\n    val: \"findings\",\n    databases: {\n      configured: configuredDatabases.success,\n      transcripts: configuredDatabases.data?.transcripts ? \"connected\" : \"error\",\n      questions: configuredDatabases.data?.questions ? \"connected\" : \"error\",\n      reports: configuredDatabases.data?.reports ? \"connected\" : \"error\",\n    }\n  };\n\n  if (configuredDatabases.errors) {\n    healthStatus.errors = configuredDatabases.errors;\n  }\n\n  return healthStatus;\n}",[235,241,275],"967f0a1bbea1eead274703e79ce3847ab183a26a9c99cb8dbb734f145e6c44a9",{"_102":274,"_95":275,"_155":276,"_157":242,"_158":137,"_159":277,"_161":162,"_104":278,"_99":160,"_164":260,"_166":279,"_167":22,"_168":-5,"_169":-7,"_98":865,"_716":866},[235,241,275],"bb250e69116c5603542bad5c505b9ed56eb7fc85c1fa3378e3e8cef150d18f53",{"_98":868,"_712":869,"_713":877},"backend/crons/README.md",{"_102":870,"_95":247,"_155":871,"_157":872,"_158":140,"_159":142,"_161":162,"_104":873,"_99":32,"_164":177,"_166":874,"_167":22,"_168":-5,"_169":-7,"_98":875,"_716":876},"7c974648-7933-11f0-96c0-0224a6c84d84","7c9746c0-7933-11f0-96c0-0224a6c84d84","0198a99c-4a38-76d0-9aab-ae2ad2c3d413",["D",1755192248891],"# Crons\n\nScheduled task handlers.",[235,294,247],"8088836f3ce944ef26ff9c8fe0376bc0a3a4ae09e8aeac59465a5c0ccfd5e6c9",{"_102":299,"_95":247,"_155":300,"_157":295,"_158":137,"_159":301,"_161":162,"_104":302,"_99":303,"_164":177,"_166":304,"_167":22,"_168":-5,"_169":-7,"_98":878,"_716":879},[235,294,247],"d4d0e1c191b64662cd09b00923687c9f173090284ac5f0349a15c642cc0bed28",{"_98":881,"_712":882,"_713":890},"backend/routes/tasks/README.md",{"_102":883,"_95":247,"_155":884,"_157":885,"_158":140,"_159":142,"_161":162,"_104":886,"_99":32,"_164":177,"_166":887,"_167":22,"_168":-5,"_169":-7,"_98":888,"_716":889},"7c9757a0-7933-11f0-96c0-0224a6c84d84","7c97580e-7933-11f0-96c0-0224a6c84d84","0198a99c-4a38-76d0-9aab-b824b02b6756",["D",1755192248891],"# Task Routes\n\nWebhook handlers for Notion integrations.",[235,343,371,247],"2c79f55ace6ef4f3c6de345b62c9874231bd71bf16b368ae9dd2a90eb5fe79c3",{"_102":376,"_95":247,"_155":377,"_157":372,"_158":137,"_159":268,"_161":162,"_104":378,"_99":379,"_164":177,"_166":380,"_167":22,"_168":-5,"_169":-7,"_98":891,"_716":892},[235,343,371,247],"29219de1325ec60d53e58fabd19da0a0ded31f48c67b6c01c03b592cf8b004be",{"_98":894,"_712":895,"_713":902},"backend/routes/tasks/_tasks.routes.ts",{"_102":896,"_95":385,"_155":897,"_157":885,"_158":140,"_159":142,"_161":162,"_104":898,"_99":32,"_164":260,"_166":899,"_167":22,"_168":-5,"_169":-7,"_98":900,"_716":901},"7c9755de-7933-11f0-96c0-0224a6c84d84","7c975642-7933-11f0-96c0-0224a6c84d84",["D",1755192248891],"import { Hono } from \"npm:hono@3.12.12\";\n\nconst app = new Hono();\n\n// Debug endpoint to check webhook configuration\napp.get(\"/debug-webhook\", (c) => {\n  const secret = Deno.env.get(\"NOTION_WEBHOOK_SECRET\");\n  return c.json({\n    hasSecret: !!secret,\n    secretLength: secret?.length || 0,\n  });\n});\n\n// Test endpoint for webhook authentication\napp.post(\"/test\", async (c) => {\n  // At this point, webhook auth has already passed\n  return c.json({ \n    success: true, \n    message: \"Webhook authentication successful\",\n    timestamp: new Date().toISOString()\n  });\n});\n\n// Main Notion webhook endpoint\napp.post(\"/notion-webhook\", async (c) => {\n  // At this point, webhook auth has already passed\n  const body = await c.req.json();\n  console.log(\"Notion webhook received:\", body);\n\n  // Process the webhook...\n  // TODO: Add webhook processing logic here\n\n  return c.json({ success: true });\n});\n\n// Catch-all handler for undefined webhook endpoints\n// This must be the LAST route defined to catch unmatched paths\napp.post(\"*\", (c) => {\n  const path = c.req.path;\n  const method = c.req.method;\n  const returnObj = {\n    error: \"Endpoint not found\",\n    path: path,\n    method: method,\n    availableEndpoints: [\n      \"GET /tasks/debug-webhook\",\n      \"POST /tasks/test\",\n      \"POST /tasks/notion-webhook\",\n    ],\n  };\n  console.log(returnObj);\n\n  return c.json(returnObj, 404);\n});\n\nexport default app;",[235,343,371,385],"a6194861b3ca44c6f899d0747e5794874d35235fb024ad901cb6057226653754",{"_102":384,"_95":385,"_155":386,"_157":372,"_158":137,"_159":268,"_161":162,"_104":387,"_99":379,"_164":260,"_166":388,"_167":22,"_168":-5,"_169":-7,"_98":903,"_716":904},[235,343,371,385],"8aa79d18620f5acd651c144145284e99690d4dd32808cd352431bfd514e586cc",{"_98":906,"_712":907,"_713":915},"backend/services/README.md",{"_102":908,"_95":247,"_155":909,"_157":910,"_158":140,"_159":142,"_161":162,"_104":911,"_99":32,"_164":177,"_166":912,"_167":22,"_168":-5,"_169":-7,"_98":913,"_716":914},"7c977410-7933-11f0-96c0-0224a6c84d84","7c977492-7933-11f0-96c0-0224a6c84d84","0198a99c-4a38-76d0-9aab-c07ccefc3c26",["D",1755192248891],"# Services\n\nExternal API integrations - pure API calls only. Services should:\n\n- Make direct API calls to external systems\n- Return consistent `{ success: boolean, data?: any, error?: string }` format\n- NOT handle environment variables or business logic\n- NOT orchestrate multiple calls",[235,448,247],"8301eb3c9b5e2f7355f26d061cc6f6083794ebb940720063fe2d3a6175cd8054",{"_102":453,"_95":247,"_155":454,"_157":449,"_158":137,"_159":218,"_161":162,"_104":455,"_99":303,"_164":177,"_166":456,"_167":22,"_168":-5,"_169":-7,"_98":916,"_716":917},[235,448,247],"9aae4e0246f55bb9415f06c85c2b0109fd95da8f2d9b369eecf15f3f380b7c38",{"_98":919,"_712":920,"_713":927},"backend/services/notion.service.ts",{"_102":921,"_95":470,"_155":922,"_157":910,"_158":140,"_159":142,"_161":162,"_104":923,"_99":32,"_164":260,"_166":924,"_167":22,"_168":-5,"_169":-7,"_98":925,"_716":926},"7c976a56-7933-11f0-96c0-0224a6c84d84","7c976ad8-7933-11f0-96c0-0224a6c84d84",["D",1755192248891],"import { Client } from \"npm:@notionhq/client\";\n\nconst notion = new Client({ auth: Deno.env.get(\"NOTION_API_KEY\") });\n\nexport async function getDatabases() {\n  try {\n    const response = await notion.search({\n      filter: {\n        property: \"object\",\n        value: \"database\",\n      },\n    });\n    return {\n      success: true,\n      data: response,\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}\n\nexport async function getDatabaseById(databaseId: string) {\n  try {\n    const response = await notion.databases.retrieve({\n      database_id: databaseId,\n    });\n    return {\n      success: true,\n      data: response,\n      timestamp: new Date().toISOString(),\n    };\n  } catch (error) {\n    return {\n      success: false,\n      error: error.message,\n      timestamp: new Date().toISOString(),\n    };\n  }\n}",[235,448,470],"8040910625120bdee1210765f9cd0c9ee0950d2d7da3540b894b417e6c41f742",{"_102":469,"_95":470,"_155":471,"_157":449,"_158":137,"_159":218,"_161":162,"_104":472,"_99":227,"_164":260,"_166":473,"_167":22,"_168":-5,"_169":-7,"_98":928,"_716":929},[235,448,470],"6dc14c55f8ecc016c4d5b4f4f0e246db362d41aec08fe8ea4642c22685aaa8bd",{"_98":931,"_712":932,"_713":940},"backend/types/README.md",{"_102":933,"_95":247,"_155":934,"_157":935,"_158":140,"_159":142,"_161":162,"_104":936,"_99":32,"_164":177,"_166":937,"_167":22,"_168":-5,"_169":-7,"_98":938,"_716":939},"7c9775c8-7933-11f0-96c0-0224a6c84d84","7c977640-7933-11f0-96c0-0224a6c84d84","0198a99c-4a38-76d0-9aab-c616707b11c3",["D",1755192248891],"# Types\n\nShared TypeScript type definitions.",[235,488,247],"ffdda641e6d6d8b32b8dfb4b81b0896b7fad006cc733aea5d69aeb8e6c5ceb85",{"_102":493,"_95":247,"_155":494,"_157":489,"_158":137,"_159":218,"_161":162,"_104":495,"_99":496,"_164":177,"_166":497,"_167":22,"_168":-5,"_169":-7,"_98":941,"_716":942},[235,488,247],"a3e3b57f3b009e3a3510f1a566f9ddd333e9c5c03aaa968daca1c6188b033efe",{"_98":944,"_712":945,"_713":953},"backend/utils/README.md",{"_102":946,"_95":247,"_155":947,"_157":948,"_158":140,"_159":142,"_161":162,"_104":949,"_99":32,"_164":177,"_166":950,"_167":22,"_168":-5,"_169":-7,"_98":951,"_716":952},"7c97773a-7933-11f0-96c0-0224a6c84d84","7c9777a8-7933-11f0-96c0-0224a6c84d84","0198a99c-4a38-76d0-9aab-cb92cb152550",["D",1755192248891],"# Utils\n\nCommon utility functions.",[235,527,247],"598ec076be7eb7bf9a25c7e67f731eda4902a463ac15acee18780a55e36be576",{"_102":532,"_95":247,"_155":533,"_157":528,"_158":137,"_159":249,"_161":162,"_104":534,"_99":287,"_164":177,"_166":535,"_167":22,"_168":-5,"_169":-7,"_98":954,"_716":955},[235,527,247],"005b52e548dabeafafb3bc9284ac93c8fb094f3233d3c9a026148c66caa785f9","hasChanges","sourceBranch","targetBranch",{"_102":140,"_104":960,"_689":961,"_95":43,"_138":145,"_99":142,"_139":-5,"_141":-5,"_691":692,"_106":-5,"_149":22,"_108":109,"_148":94,"_16":118,"_147":682},["D",1755176650069],["D",1755192248891],"pull","actionData","errors"]
lightweight

findings

Mergefindings/mainintofindings-base/main

_townie/00-setup.md+1-1

backend/README.md+2-1

backend/controllers/README.md+17-1

backend/controllers/health.controller.ts+12-10

backend/crons/README.md+27-1

backend/routes/tasks/README.md+99-1

backend/routes/tasks/_tasks.routes.ts+139-0

backend/services/README.md+71-5

backend/services/notion.service.ts+657-0

backend/types/README.md+10-1

backend/utils/README.md+14-1

_townie/2-stage-questions-openai.md+496-0

_townie/email-PDF-attachments.md+802-0

_townie/pdf-to-blob.md+537-0

backend/controllers/emails.controller.ts+287-0

backend/controllers/findings.controller.ts+1112-0

backend/controllers/tasks.controller.ts+1290-0

Cbackend/crons/findings-analysis.ts+45-0

backend/emails

backend/emails/README.md+80-0

Ebackend/emails/pdf-attachments.ts+218-0

backend/services/email.service.ts+206-0

backend/services/openai.service.ts+401-0

backend/types/email.types.ts+47-0

backend/types/findings.types.ts+118-0

backend/types/notion.types.ts+42-0

backend/utils/email.utils.ts+29-0

backend/utils/notion-properties.utils.ts+214-0

backend/utils/pdfExtractor.ts+68-0

backend/utils/process-timing.utils.ts+123-0

backend/utils/text-redaction.utils.ts+50-0
