403 Forbidden

\", 403);\n\n const [\n allUsers,\n totalUsersResult,\n totalExecsResult,\n totalCostResult,\n costPerUser,\n pendingVerifResult,\n pendingTasksResult,\n recentExecutions,\n ] = await Promise.all([\n sqlite.execute(`SELECT * FROM users ORDER BY created_at DESC`),\n sqlite.execute(`SELECT COUNT(*) as count FROM users`),\n sqlite.execute(`SELECT COUNT(*) as count FROM sandbox_executions`),\n sqlite.execute(`SELECT COALESCE(SUM(cost_usd), 0) as total FROM sandbox_executions`),\n sqlite.execute(`\n SELECT u.email, u.name,\n COUNT(se.id) as execution_count,\n COALESCE(SUM(se.cost_usd), 0) as total_cost\n FROM users u\n LEFT JOIN sandbox_executions se ON se.user_id = u.id\n GROUP BY u.id\n ORDER BY total_cost DESC\n `),\n sqlite.execute(`SELECT COUNT(*) as count FROM users WHERE status = 'pending_verification'`),\n sqlite.execute(`SELECT COUNT(*) as count FROM scheduled_tasks WHERE status = 'pending'`),\n sqlite.execute(`\n SELECT se.*, u.email as user_email\n FROM sandbox_executions se\n JOIN users u ON se.user_id = u.id\n ORDER BY se.created_at DESC\n LIMIT 100\n `),\n ]);\n\n return htmlResponse(\n adminDashboardPage({\n currentUser: user,\n allUsers: allUsers.rows as Record[],\n recentExecutions: recentExecutions.rows as Record[],\n totalUsers: (totalUsersResult.rows[0] as Record).count as number,\n totalExecutions: (totalExecsResult.rows[0] as Record).count as number,\n totalCost: (totalCostResult.rows[0] as Record).total as number,\n costPerUser: costPerUser.rows as Record[],\n pendingVerifications: (pendingVerifResult.rows[0] as Record).count as number,\n pendingTasks: (pendingTasksResult.rows[0] as Record).count as number,\n }),\n );\n }\n\n if (url.pathname.startsWith(\"/admin/user/\")) {\n const user = await getSessionUser(req);\n if (!user) return redirect(\"/login\");\n if (!user.is_admin) return htmlResponse(\"

403 Forbidden

\", 403);\n\n const targetEmail = decodeURIComponent(url.pathname.replace(\"/admin/user/\", \"\"));\n\n const targetUserResult = await sqlite.execute({\n sql: `SELECT * FROM users WHERE email = ?`,\n args: [targetEmail],\n });\n\n if (targetUserResult.rows.length === 0) {\n return htmlResponse(\"

User not found

\", 404);\n }\n\n const targetUser = targetUserResult.rows[0] as Record;\n const targetUserId = targetUser.id as number;\n\n const [executions, scheduledTasks, totalExecResult, totalCostResult, pendingEmailResult] =\n await Promise.all([\n sqlite.execute({\n sql: `SELECT * FROM sandbox_executions WHERE user_id = ? ORDER BY created_at DESC LIMIT 50`,\n args: [targetUserId],\n }),\n sqlite.execute({\n sql: `SELECT * FROM scheduled_tasks WHERE user_id = ? ORDER BY CASE WHEN status = 'pending' THEN 0 ELSE 1 END, next_run_at DESC LIMIT 50`,\n args: [targetUserId],\n }),\n sqlite.execute({\n sql: `SELECT COUNT(*) as count FROM sandbox_executions WHERE user_id = ?`,\n args: [targetUserId],\n }),\n sqlite.execute({\n sql: `SELECT COALESCE(SUM(cost_usd), 0) as total FROM sandbox_executions WHERE user_id = ?`,\n args: [targetUserId],\n }),\n sqlite.execute({\n sql: `SELECT COUNT(*) as count FROM pending_emails WHERE user_id = ?`,\n args: [targetUserId],\n }),\n ]);\n\n return htmlResponse(\n adminUserDetailPage({\n currentUser: user,\n targetUser,\n executions: executions.rows as Record[],\n scheduledTasks: scheduledTasks.rows as Record[],\n totalExecutions: (totalExecResult.rows[0] as Record).count as number,\n totalCost: (totalCostResult.rows[0] as Record).total as number,\n pendingEmails: (pendingEmailResult.rows[0] as Record).count as number,\n }),\n );\n }\n\n // ============ JSON API ============\n\n if (url.pathname === \"/api/stats\") {\n const [users, scheduledTasks, executions] = await Promise.all([\n sqlite.execute(\"SELECT COUNT(*) as count FROM users\"),\n sqlite.execute(\n \"SELECT status, COUNT(*) as count FROM scheduled_tasks GROUP BY status ORDER BY count DESC\",\n ),\n sqlite.execute(\"SELECT COUNT(*) as count FROM sandbox_executions\"),\n ]);\n\n const tasksByStatus: Record = {};\n for (const row of scheduledTasks.rows) {\n const r = row as Record;\n tasksByStatus[r.status as string] = r.count as number;\n }\n\n return jsonResponse({\n name: \"Email Agent\",\n version: \"v2 (sandbox architecture)\",\n status: \"running\",\n stats: {\n users: (users.rows[0] as Record).count,\n total_executions: (executions.rows[0] as Record).count,\n scheduled_tasks_by_status: tasksByStatus,\n },\n email: \"memory-do@valtown.email\",\n });\n }\n\n // 404\n return htmlResponse(\n `

404

Page not found.

← Home\n

`,\n 404,\n );\n}\n","memory-do",[149],"c2a4fa44-19b1-11f1-931b-42dde27851f2","email-handler-v2.ts","5e843112-18bf-11f1-867c-42dde27851f2",103,["D",1772838669058],7,"email","/**\n * Email Handler v2 — Sandbox-based architecture with email verification.\n *\n * When an email arrives:\n * 1. Identify/create the user from the sender's email\n * 2. If the user is NEW (pending_verification):\n * a. Store their email in `pending_emails`\n * b. Send a verification email with a magic link + welcome instructions\n * c. Stop processing — their email will be forwarded after they verify\n * 3. If the user is UNVERIFIED (pending_verification, not first email):\n * a. Store the email in `pending_emails`\n * b. Send a gentle nudge to verify\n * c. Stop processing\n * 4. If the user is VERIFIED:\n * a. Write the email to the sandbox inbox\n * b. Boot a sandbox and run agent-loop.mjs\n * c. Read the agent's response from the outbox\n * d. Process any schedule requests\n * e. Email the response back\n */\n\nimport { email as sendEmail } from \"https://esm.town/v/std/email\";\nimport { sqlite } from \"https://esm.town/v/std/sqlite/main.ts\";\nimport { marked } from \"npm:marked@15\";\nimport {\n initDatabaseV2,\n getOrCreateUserV2,\n addScheduledTask,\n storePendingEmail,\n createEmailVerificationToken,\n type UserV2,\n} from \"./db/schema-v2.ts\";\nimport {\n executeSandboxTask,\n type ScheduleRequest,\n type FileAttachment,\n} from \"./lib/sandbox-orchestrator.ts\";\nimport { parseSchedule } from \"./lib/scheduler.ts\";\n\n// Email type from Val Town (with attachments)\ninterface Email {\n from: string;\n to: string[];\n subject: string | null;\n text: string | null;\n html: string | null;\n attachments: {\n name: string;\n type: string;\n size: number;\n text: () => Promise;\n arrayBuffer: () => Promise;\n }[];\n}\n\n/**\n * Generate a fingerprint for deduplication.\n * Uses sender + subject + body prefix + timestamp (rounded to 60s window).\n */\nfunction emailFingerprint(email: Email): string {\n const bodySnippet = (email.text || email.html || \"\").slice(0, 200);\n const raw = `${email.from}|${email.subject || \"\"}|${bodySnippet}`;\n // Simple hash — we don't need crypto-strength, just uniqueness\n let hash = 0;\n for (let i = 0; i < raw.length; i++) {\n const ch = raw.charCodeAt(i);\n hash = ((hash << 5) - hash + ch) | 0;\n }\n return `email-${Math.abs(hash).toString(36)}`;\n}\n\n/**\n * Ensure the dedup table exists and check/claim a fingerprint.\n * Returns true if this email should be processed (first time seen),\n * false if it's a duplicate.\n */\nasync function claimEmailProcessing(fingerprint: string): Promise {\n await sqlite.execute(`\n CREATE TABLE IF NOT EXISTS email_dedup (\n fingerprint TEXT PRIMARY KEY,\n created_at TEXT DEFAULT (datetime('now')),\n status TEXT DEFAULT 'processing'\n )\n `);\n\n // Clean up old entries (older than 1 hour)\n await sqlite.execute(\n `DELETE FROM email_dedup WHERE created_at < datetime('now', '-1 hour')`,\n );\n\n // Try to insert — if fingerprint already exists, this is a duplicate\n try {\n await sqlite.execute({\n sql: `INSERT INTO email_dedup (fingerprint) VALUES (?)`,\n args: [fingerprint],\n });\n return true; // We claimed it\n } catch (err) {\n const errMsg = err instanceof Error ? err.message : String(err);\n if (errMsg.includes(\"UNIQUE\") || errMsg.includes(\"constraint\")) {\n return false; // Duplicate\n }\n // Some other error — let it through to be safe\n console.warn(`⚠️ Dedup check error (allowing processing): ${errMsg}`);\n return true;\n }\n}\n\n/**\n * Build the task description from the email (shared between immediate processing\n * and stored pending emails).\n */\nasync function buildTaskDescription(email: Email): Promise {\n const body = email.text || email.html || \"(empty email)\";\n const subject = email.subject || \"Your message\";\n\n let taskDescription = `## Email from: ${email.from}\\n`;\n taskDescription += `**Subject**: ${subject}\\n\\n`;\n taskDescription += body;\n\n // Process text-readable attachments into the description\n if (email.attachments && email.attachments.length > 0) {\n console.log(`📎 Processing ${email.attachments.length} attachment(s)...`);\n taskDescription += \"\\n\\n## Attachments\\n\";\n\n for (const file of email.attachments) {\n try {\n const textContent = await file.text();\n if (textContent && textContent.length > 0) {\n const truncated =\n textContent.length > 10000\n ? textContent.slice(0, 10000) + \"\\n\\n[...truncated...]\"\n : textContent;\n taskDescription += `\\n### ${file.name} (${file.type})\\n\\`\\`\\`\\n${truncated}\\n\\`\\`\\`\\n`;\n } else {\n taskDescription += `\\n### ${file.name} (${file.type}, ${file.size} bytes) — binary file, content not shown\\n`;\n }\n } catch {\n taskDescription += `\\n### ${file.name} (${file.type}, ${file.size} bytes) — could not read content\\n`;\n }\n }\n }\n\n return taskDescription;\n}\n\n/**\n * Serialize attachments to JSON for storage in pending_emails.\n */\nasync function serializeAttachments(\n email: Email,\n): Promise {\n if (!email.attachments || email.attachments.length === 0) return undefined;\n\n const attachments = [];\n for (const file of email.attachments) {\n try {\n const textContent = await file.text();\n attachments.push({\n name: file.name,\n type: file.type,\n size: file.size,\n textContent: textContent?.slice(0, 10000) || null,\n });\n } catch {\n attachments.push({\n name: file.name,\n type: file.type,\n size: file.size,\n textContent: null,\n });\n }\n }\n return JSON.stringify(attachments);\n}\n\n// ============ VERIFICATION EMAIL ============\n\nconst VERIFY_BASE_URL = \"https://memory-do.val.run\";\n\nasync function sendVerificationEmail(\n user: UserV2,\n isFirstEmail: boolean,\n) {\n const token = await createEmailVerificationToken(user.id);\n const verifyUrl = `${VERIFY_BASE_URL}/verify-email?token=${token}`;\n\n if (isFirstEmail) {\n // Welcome email with verification + instructions\n await sendEmail({\n to: user.email,\n subject: \"Welcome to Email Agent — please verify your email\",\n replyTo: \"memory-do@valtown.email\",\n text: `Welcome to Email Agent! 📧\n\nBefore I can process your emails, I need to verify that this is really you.\n\nClick here to verify your email address:\n${verifyUrl}\n\nThis link expires in 24 hours.\n\n────────────────────────────\n\nOnce verified, here's what you can do — just send an email to memory-do@valtown.email:\n\n📝 Remember things\n \"I work at Acme Corp\"\n \"My friend Sarah is a designer\"\n\n✅ Manage tasks & reminders\n \"Remind me Thursday to call the dentist\"\n \"Every Monday, summarize my active tasks\"\n\n🔍 Research & analysis\n \"What's the latest news on WebAssembly?\"\n \"Summarize this article: [paste URL]\"\n\n💻 Run code & work with files\n Send code snippets, CSVs, or data — I can analyze them\n\n🎨 Personalize me\n \"Call me Paul\"\n \"Be more casual and use emojis\"\n I'll remember your preferences across conversations.\n\nYour first email has been saved and will be processed as soon as you verify.\n\n— Email Agent`,\n html: `\n\n\n\n

📧 Welcome to Email Agent!

Before I can process your emails, I need to verify that this is really you.

\n\n

\n Verify My Email Address\n

\n\n

This link expires in 24 hours.

\n\n

Once verified, here's what you can do — just send an email to memory-do@valtown.email:

\n\n \n \n \n \n \n \n

📝	Remember things \"I work at Acme Corp\" · \"My friend Sarah is a designer\"
✅	Manage tasks & reminders \"Remind me Thursday to call the dentist\" · \"Every Monday, summarize my tasks\"
🔍	Research & analysis \"What's the latest news on WebAssembly?\"
💻	Run code & work with files Send code, CSVs, or data — I can analyze them
🎨	Personalize me \"Call me Paul\" · \"Be more casual\" — I remember your preferences

\n\n

\n Your first email has been saved and will be processed as soon as you verify. 🎉\n

\n\n`,\n });\n } else {\n // Gentle nudge — they've sent another email but haven't verified yet\n await sendEmail({\n to: user.email,\n subject: \"Email Agent — please verify to process your emails\",\n replyTo: \"memory-do@valtown.email\",\n text: `Hi! I've saved your email, but I still need you to verify your email address before I can process anything.\n\nClick here to verify: ${verifyUrl}\n\nThis link expires in 24 hours. Once you verify, all your saved emails will be processed.\n\n— Email Agent`,\n html: `\n\n\n\n

📧 Email Agent

I've saved your email, but I still need you to verify your email address before I can process anything.

\n Verify My Email\n

This link expires in 24 hours. Once you verify, all your saved emails will be processed. 🎉

\n\n`,\n });\n }\n}\n\n// ============ MAIN HANDLER ============\n\nexport default async function (email: Email) {\n console.log(\n `📧 Email received from: ${email.from}, subject: ${email.subject}`,\n );\n\n // Guard against self-send loops — ignore emails from Val Town addresses\n if (email.from.includes(\"@valtown.email\")) {\n console.log(\n \"🚫 Ignoring email from @valtown.email (self-send loop prevention)\",\n );\n return;\n }\n\n // Deduplication — prevent the same email from being processed twice\n const fingerprint = emailFingerprint(email);\n const shouldProcess = await claimEmailProcessing(fingerprint);\n if (!shouldProcess) {\n console.log(\n `🔁 Duplicate email detected (fingerprint: ${fingerprint}), skipping`,\n );\n return;\n }\n console.log(\n `🆔 Email fingerprint: ${fingerprint} (first seen, processing)`,\n );\n\n await initDatabaseV2();\n\n // 1. Get or create user\n const senderEmail = email.from;\n const { user, isNew } = await getOrCreateUserV2(senderEmail);\n console.log(\n `👤 User: ${user.email} (id: ${user.id}, status: ${user.status}, isNew: ${isNew})`,\n );\n\n // 2. Check if user needs verification\n if (user.status === \"pending_verification\") {\n console.log(\n `🔒 User is unverified — storing email and sending ${isNew ? \"welcome\" : \"reminder\"} verification email`,\n );\n\n // Build and store the email body + attachments for later\n const body = email.text || email.html || \"(empty email)\";\n const attachmentsJson = await serializeAttachments(email);\n\n await storePendingEmail({\n userId: user.id,\n subject: email.subject,\n body,\n attachmentsJson,\n });\n\n // Send verification email\n try {\n await sendVerificationEmail(user, isNew);\n console.log(`📤 Verification email sent to ${user.email}`);\n } catch (err) {\n console.error(\"Failed to send verification email:\", err);\n }\n\n return;\n }\n\n // 3. User is verified — process normally\n await processVerifiedEmail(user, email);\n}\n\n/**\n * Process an email for a verified user — the normal happy path.\n */\nasync function processVerifiedEmail(user: UserV2, email: Email) {\n const subject = email.subject || \"Your message\";\n const taskDescription = await buildTaskDescription(email);\n\n // Execute in the user's sandbox\n console.log(\"🚀 Launching sandbox for task execution...\");\n const result = await executeSandboxTask(\n user,\n {\n title: subject,\n description: taskDescription,\n },\n { type: \"email\" },\n );\n\n console.log(\n `📊 Sandbox result: success=${result.success}, turns=${result.turns}, elapsed=${result.elapsedMs}ms, cost=$${result.costUsd?.toFixed(4) ?? \"unknown\"}`,\n );\n\n // Process schedule requests from the agent\n if (result.schedules.length > 0) {\n console.log(`⏰ Agent requested ${result.schedules.length} schedule(s)`);\n for (const schedule of result.schedules) {\n try {\n await processScheduleRequest(user.id, schedule);\n console.log(\n `⏰ Scheduled: \"${schedule.title}\" → ${schedule.schedule}`,\n );\n } catch (err) {\n console.error(`Failed to schedule \"${schedule.title}\":`, err);\n }\n }\n }\n\n // Send reply email\n if (result.response) {\n try {\n const htmlBody = await marked.parse(result.response);\n const htmlWrapped = `\n\n\n\n\n${htmlBody}\n\n`;\n\n const emailOpts: Record = {\n to: user.email,\n subject: `Re: ${subject}`,\n text: result.response,\n html: htmlWrapped,\n replyTo: \"memory-do@valtown.email\",\n };\n\n // Attach files the agent created\n if (result.fileAttachments.length > 0) {\n const attachments = result.fileAttachments.map(\n (file: FileAttachment) => ({\n content: file.content,\n filename: file.filename,\n type: file.mimeType,\n disposition: \"attachment\" as const,\n }),\n );\n emailOpts.attachments = attachments;\n console.log(\n `📎 Attaching ${attachments.length} file(s): ${attachments.map((a: { filename: string }) => a.filename).join(\", \")}`,\n );\n }\n\n await sendEmail(emailOpts as Parameters[0]);\n console.log(\n `📤 Reply sent to ${user.email} (text + html${result.fileAttachments.length > 0 ? ` + ${result.fileAttachments.length} attachment(s)` : \"\"})`,\n );\n } catch (error) {\n console.error(\"Failed to send reply:\", error);\n }\n }\n\n console.log(\"✅ Email processing complete\");\n}\n\n/**\n * Process pending emails for a user who just verified.\n * Called from main.ts after a successful email verification.\n * Replays all stored emails through the sandbox.\n */\nexport async function processPendingEmailsForUser(user: UserV2) {\n const { getPendingEmails, deletePendingEmails } = await import(\n \"./db/schema-v2.ts\"\n );\n const pendingEmails = await getPendingEmails(user.id);\n\n if (pendingEmails.length === 0) {\n console.log(`📭 No pending emails for user ${user.email}`);\n return;\n }\n\n console.log(\n `📬 Processing ${pendingEmails.length} pending email(s) for ${user.email}`,\n );\n\n // Process each pending email as a sandbox task\n for (const pending of pendingEmails) {\n const subject = pending.subject || \"Your message\";\n\n // Rebuild the task description from stored data\n let taskDescription = `## Email from: ${user.email}\\n`;\n taskDescription += `**Subject**: ${subject}\\n\\n`;\n taskDescription += pending.body;\n\n // Re-attach attachment info if we stored it\n if (pending.attachments_json) {\n try {\n const attachments = JSON.parse(pending.attachments_json);\n if (attachments.length > 0) {\n taskDescription += \"\\n\\n## Attachments\\n\";\n for (const att of attachments) {\n if (att.textContent) {\n taskDescription += `\\n### ${att.name} (${att.type})\\n\\`\\`\\`\\n${att.textContent}\\n\\`\\`\\`\\n`;\n } else {\n taskDescription += `\\n### ${att.name} (${att.type}, ${att.size} bytes) — binary file, content not shown\\n`;\n }\n }\n }\n } catch {\n // Ignore malformed JSON\n }\n }\n\n console.log(`🚀 Processing pending email: \"${subject}\"...`);\n const result = await executeSandboxTask(\n user,\n { title: subject, description: taskDescription },\n { type: \"email\" },\n );\n\n console.log(\n `📊 Pending email result: success=${result.success}, turns=${result.turns}`,\n );\n\n // Process schedules\n if (result.schedules.length > 0) {\n for (const schedule of result.schedules) {\n try {\n await processScheduleRequest(user.id, schedule);\n } catch (err) {\n console.error(`Failed to schedule \"${schedule.title}\":`, err);\n }\n }\n }\n\n // Send reply\n if (result.response) {\n try {\n const htmlBody = await marked.parse(result.response);\n const htmlWrapped = `\n\n\n\n\n${htmlBody}\n\n`;\n\n await sendEmail({\n to: user.email,\n subject: `Re: ${subject}`,\n text: result.response,\n html: htmlWrapped,\n replyTo: \"memory-do@valtown.email\",\n });\n console.log(`📤 Pending email reply sent for \"${subject}\"`);\n } catch (error) {\n console.error(\"Failed to send pending email reply:\", error);\n }\n }\n }\n\n // Clean up — all pending emails have been processed\n await deletePendingEmails(user.id);\n console.log(\n `🧹 Cleared ${pendingEmails.length} pending email(s) for ${user.email}`,\n );\n}\n\n/**\n * Process a schedule request from the sandbox agent.\n */\nasync function processScheduleRequest(\n userId: number,\n schedule: ScheduleRequest,\n) {\n const parsed = await parseSchedule(schedule.schedule);\n\n await addScheduledTask({\n userId,\n title: schedule.title,\n description: schedule.description,\n nextRunAt: parsed.nextRunAt,\n scheduleDescription: schedule.schedule,\n isRecurring: schedule.is_recurring,\n });\n}\n",[158],"cf56f678-19ae-11f1-ad6c-42dde27851f2","README.md","bb517fd8-181c-11f1-a52e-42dde27851f2",81,["D",1772837401848],1,"# 📧 Email Agent\n\nAn email-based personal AI agent built on [Val Town](https://val.town). Interact entirely via email — send messages, tasks, ideas, and memories to **`memory-do@valtown.email`** and the agent processes everything inside a persistent per-user sandbox.\n\n## How It Works\n\nSend an email. The agent reads it, figures out what you mean, and acts on it.\n\n```\nYou → \"Remind me Thursday to call the dentist\"\n → \"I like chickens\"\n → \"What's the latest news on WebAssembly?\"\n → \"Every Monday, give me a summary of my active tasks\"\n```\n\nUnder the hood:\n\n1. **Receive** — Your email arrives and is written to your personal sandbox inbox\n2. **Boot** — A Deno sandbox spins up with your persistent volume mounted\n3. **Execute** — Claude reads your `CLAUDE.md`, checks the inbox, and handles everything using full Linux tooling (Bash, file I/O, git, curl, etc.)\n4. **Reply** — You get an email back with results, follow-up questions, and any generated file attachments (rendered as HTML with a plain-text fallback)\n\n### Voice-Friendly\n\nEmails sent via voice transcription work great. The agent is tuned for misspellings, run-on sentences, and implicit intent — it interprets what you *mean*, not what you literally typed.\n\n## Sandbox Architecture\n\nEach user gets their own **persistent Deno sandbox volume** — a private filesystem that acts as the agent's brain. There's no classifier step and no custom tool-use loop. The agent runs inside the sandbox using the **Claude Agent SDK** with native Linux tools.\n\n```\n┌───────────────────────────────────────────────────────┐\n│ VAL TOWN (orchestrator) │\n│ │\n│ email-handler-v2.ts → receives email │\n│ ├─ Get/create user │\n│ ├─ Write email to sandbox inbox │\n│ ├─ Boot sandbox → run agent-loop.mjs │\n│ ├─ Read response from sandbox outbox │\n│ ├─ Process schedule.json → scheduled_tasks │\n│ └─ Send reply email (HTML + plain text) │\n│ │\n│ task-runner-v2.ts → runs every 15 min │\n│ ├─ Query due scheduled_tasks │\n│ ├─ Boot sandbox for each due task │\n│ └─ Email results, reschedule if recurring │\n└───────────────────┬───────────────────────────────────┘\n │\n ▼\n┌───────────────────────────────────────────────────────┐\n│ DENO SANDBOX (per-user microVM) │\n│ │\n│ Runs: node /opt/agent/agent-loop.mjs │\n│ Uses: Claude Agent SDK with Bash/Read/Write/Edit/ │\n│ Glob/Grep tools │\n│ Mounts: user's persistent volume at /home/user/ │\n│ │\n│ /home/user/ │\n│ ├── CLAUDE.md ← agent instructions │\n│ ├── agents.md ← persona/identity │\n│ ├── TODO.md ← persistent task list │\n│ ├── memories/ ← knowledge as markdown files │\n│ ├── ideas/ ← captured ideas │\n│ ├── people/ ← people directory │\n│ ├── repos/ ← cloned git repositories │\n│ ├── workspace/ ← scratch area │\n│ ├── .ssh/ ← SSH keys │\n│ └── .mail/ │\n│ ├── inbox/ ← tasks arrive here │\n│ └── outbox/ ← responses go here │\n└───────────────────────────────────────────────────────┘\n```\n\n### The Volume IS the User's Brain\n\nAll user state lives on their persistent sandbox volume as plain files:\n\n| What | Where |\n|---|---|\n| Memories & facts | `memories/*.md` (one file per topic, Claude organizes) |\n| Tasks & reminders | `TODO.md` (Claude maintains it directly) |\n| Ideas | `ideas/*.md` (one per idea) |\n| People | `people/*.md` (one per person) |\n| Preferences | Embedded in `CLAUDE.md` (Claude edits it!) |\n| Git repos | `repos/` (persistent across sessions) |\n| SSH keys | `.ssh/` (standard Unix location) |\n\nNo SQL for content storage. Claude manages files naturally because it has full filesystem access.\n\n### Available Tools (in the sandbox)\n\nThe agent uses the Claude Agent SDK's built-in tools — not custom function calls:\n\n| Tool | What It Does |\n|---|---|\n| `Bash` | Run any command — git, npm, python, curl, etc. |\n| `Read` | Read any file on the volume |\n| `Write` | Write any file on the volume |\n| `Edit` | Edit files with diffs |\n| `Glob` | Find files by pattern |\n| `Grep` | Search file contents |\n\nThis means the agent can do multi-step work: search your memories with grep, fetch a web page with curl, cross-reference with your stored projects, save what it learned as a markdown file, clone a repo, run code — all in one email exchange.\n\n## Task Types\n\n| Type | Behavior | Example |\n|---|---|---|\n| **Instant** | Executes immediately, reply in same email | \"What's the weather in London?\" |\n| **Future** | Scheduled for a specific time | \"Remind me Thursday at 10am to call dentist\" |\n| **Recurring** | Repeats on a schedule | \"Every Monday, summarize my active tasks\" |\n\nFuture and recurring tasks are processed by a task runner that checks every 15 minutes. The agent writes a `schedule.json` to its outbox, and the orchestrator picks it up and inserts it into the `scheduled_tasks` table.\n\n## Per-User Personalization\n\nEach user's sandbox has a `CLAUDE.md` and `agents.md` that the agent reads at the start of every session. The agent can **edit these files directly** — so when you say:\n\n- *\"Be more casual and use emojis\"*\n- *\"Call me Paul\"*\n- *\"Always use bullet points\"*\n\n…the agent updates its own instructions. Changes persist across sessions because the volume is persistent.\n\n## Attachments\n\nFull bidirectional support:\n\n- **Inbound** — Email attachments (CSV, JSON, text, code files, etc.) are included in the task description sent to the sandbox\n- **Outbound** — The agent can generate files inside the sandbox, which are read from the outbox and sent back as email attachments\n\n## Project Structure\n\n```\nemail-agent/\n README.md ← You are here\n CLAUDE.md — Internal dev guide and conventions\n main.ts — HTTP admin/status endpoint\n email-handler-v2.ts — Email ingestion → sandbox execution\n task-runner-v2.ts — Scheduled task processor (every 15 min)\n lib/\n sandbox-orchestrator.ts — Volume CRUD, sandbox boot, result reading\n sandbox-agent-loop.ts — Source for agent-loop.mjs (runs in sandbox)\n sandbox-claude-md.ts — Builds CLAUDE.md and agents.md for volumes\n scheduler.ts — Natural language → ISO datetime\n db/\n schema-v2.ts — SQLite schema (users, scheduled_tasks, executions)\n schema.ts — Legacy v1 schema (entries, FTS, links)\n queries.ts — Legacy v1 queries\n admin/\n build-snapshot.ts — Builds the toolchain sandbox snapshot\n test-orchestrator.ts — Test harness for sandbox orchestration\n```\n\n### Tech Stack\n\n- **Runtime**: [Val Town](https://val.town) (Deno-based serverless)\n- **Agent**: [Claude Agent SDK](https://docs.anthropic.com/en/docs/build-with-claude/tool-use) running inside Deno sandboxes\n- **Sandboxes**: Deno Sandbox SDK — per-user persistent volumes with full Linux tooling\n- **Database**: Project-scoped SQLite (thin routing layer only — users, scheduled tasks, execution logs)\n- **Email**: Val Town email sending/receiving\n- **LLM**: Claude via Anthropic SDK (agent loop + scheduler)\n\n### Database (SQLite)\n\nSQLite is a thin routing layer — it does **not** store user content (that lives on sandbox volumes):\n\n| Table | Purpose |\n|---|---|\n| `users` | Email → user mapping + volume reference |\n| `scheduled_tasks` | When to wake up a sandbox (task runner polls this) |\n| `sandbox_executions` | Execution log for debugging and cost tracking |\n| `email_dedup` | Short-lived deduplication table for email retries |\n\n## Environment Variables\n\n| Variable | Required | Description |\n|---|---|---|\n| `ANTHROPIC_API_KEY` | Yes | Used by Claude API calls (agent loop, scheduler) |\n| `DENO_DEPLOY_TOKEN` | Yes | Used for Deno sandbox and volume management |\n\n## Admin / Debugging\n\nThe HTTP endpoint (`main.ts`) provides:\n\n- `GET /` or `GET /stats` — System-wide stats (entry counts, users, links)\n- `GET /user/:email` — Full user profile with entries grouped by type\n- `GET /migrate` — Run one-time data migration from legacy tables\n\n> **Note**: `main.ts` currently queries the legacy v1 schema tables (`entries`, `entry_links`). A future update will switch it to the v2 tables.\n\n## Limitations\n\n- **Sandbox boot latency** — ~2–5 seconds per execution; not ideal for trivial questions\n- **Concurrency** — Deno Sandbox has a concurrent execution limit during pre-release; may need queuing with multiple simultaneous users\n- **No admin auth** — HTTP endpoints are open\n- **No rate limiting** — No flood protection\n- **No cost alerting** — `sandbox_executions` logs cost but there's no per-user limits or alerts yet\n- **File attachments (outbound)** — Stubbed but not yet fully implemented\n- **SSH key management** — Infrastructure exists (`.ssh/` dir on volume) but no user-facing flow yet\n- **Snapshot rollout** — No automated process for updating the sandbox toolchain across existing users\n\n## License\n\nBuilt on Val Town. Do what you want with it.\n",[167],"fdea35ba-19b1-11f1-9e9a-42dde27851f2","CLAUDE.md","5e83aa9e-18bf-11f1-867c-42dde27851f2",109,["D",1772838768498],"# Email Agent System — Sandbox Architecture\n\nAn email-based personal AI agent built on Val Town. Users interact entirely via email — sending messages, tasks, ideas, and memories to `memory-do@valtown.email`. The agent processes requests inside per-user Deno Sandboxes with full Linux tooling.\n\n## Architecture Overview\n\n```\n┌───────────────────────────────────────────────────────────────┐\n│ VAL TOWN (orchestrator) │\n│ │\n│ email-handler-v2.ts (email) → receives email │\n│ │ │\n│ ├─ Get/create user (SQLite: users table) │\n│ ├─ Write email to user's sandbox inbox │\n│ ├─ Boot sandbox → run agent-loop.mjs │\n│ ├─ Read response from sandbox outbox │\n│ ├─ Process schedule.json → scheduled_tasks table │\n│ └─ Send reply email │\n│ │\n│ task-runner-v2.ts (interval, every 15 min) │\n│ │ │\n│ ├─ Query scheduled_tasks WHERE next_run_at <= now │\n│ ├─ For each due task: │\n│ │ ├─ Write task to user's sandbox inbox │\n│ │ ├─ Boot sandbox → run agent-loop.mjs │\n│ │ ├─ Read response, email user │\n│ │ └─ Reschedule if recurring │\n│ └─ Done │\n│ │\n│ main.ts (http) → admin/status dashboard │\n└───────────────────┬───────────────────────────────────────────┘\n │\n ▼\n┌───────────────────────────────────────────────────────────────┐\n│ DENO SANDBOX (per-user microVM) │\n│ │\n│ Runs: node /opt/agent/agent-loop.mjs │\n│ Uses: Claude Agent SDK with Bash/Read/Write/Edit/Glob/Grep │\n│ Mounts: user's persistent volume at /home/user/ │\n│ │\n│ /home/user/ │\n│ ├── CLAUDE.md ← agent instructions (Claude edits it!) │\n│ ├── agents.md ← persona/identity │\n│ ├── TODO.md ← persistent task list │\n│ ├── memories/ ← knowledge as markdown files │\n│ ├── ideas/ ← captured ideas │\n│ ├── people/ ← people directory │\n│ ├── repos/ ← cloned git repositories │\n│ ├── workspace/ ← scratch area │\n│ ├── .ssh/ ← SSH keys │\n│ ├── .mail/ │\n│ │ ├── inbox/ ← tasks arrive here (written by VT) │\n│ │ └── outbox/ ← responses go here (written by agent) │\n│ └── .agent-results/ ← cycle metrics (turns, cost, timing) │\n└───────────────────────────────────────────────────────────────┘\n```\n\n## File Structure\n\n```\nemail-agent/\n CLAUDE.md — This file. System docs and conventions.\n README.md — Project overview\n\n email-handler-v2.ts — Email ingestion → verification or sandbox execution (type: email)\n task-runner-v2.ts — Scheduled task polling → sandbox execution (type: interval)\n main.ts — HTTP: home page, auth, dashboard, email verification (type: http)\n db/\n schema-v2.ts — Users, scheduled_tasks, sandbox_executions, pending_emails, auth tables\n lib/\n auth.ts — Magic-link auth for web dashboard + cleanup helpers\n pages.ts — HTML page templates (home, login, dashboard, verification)\n sandbox-orchestrator.ts — Volume CRUD, sandbox boot, result reading\n sandbox-agent-loop.ts — Source code for agent-loop.mjs (runs in sandbox)\n sandbox-claude-md.ts — Builds CLAUDE.md and agents.md for user volumes\n scheduler.ts — Natural language → ISO datetime\n admin/\n build-snapshot.ts — Builds the toolchain sandbox snapshot\n test-orchestrator.ts — Test harness for sandbox orchestration\n```\n\n## Key Design Principles\n\n### 1. The Volume IS the User's Brain\nAll user state lives on their persistent sandbox volume as plain files:\n- **Memories** → `memories/*.md` (one file per topic, Claude organizes)\n- **Tasks** → `TODO.md` (Claude maintains it directly)\n- **Ideas** → `ideas/*.md` (one per idea)\n- **People** → `people/*.md` (one per person)\n- **Preferences** → embedded in `CLAUDE.md` (Claude edits it!)\n- **SSH keys** → `.ssh/` (standard Unix location)\n- **Git repos** → `repos/` (persistent across sessions)\n\nNo more SQL for content storage. No FTS tables. No entry_links graph.\nClaude manages files naturally because it has Bash/Read/Write/Edit tools.\n\n### 2. SQLite = Thin Routing Layer\nSQLite only stores what Val Town needs to operate:\n- **users**: email → user mapping + volume reference\n- **scheduled_tasks**: when to wake up the sandbox (task runner polls this)\n- **sandbox_executions**: execution log for debugging and cost tracking\n\n### 3. The Sandbox Agent Does Everything\nNo more classifier, no more custom tool definitions, no more agent.ts loop.\nThe agent-loop.mjs inside the sandbox uses the **Claude Agent SDK** with:\n- `Bash` — run any command (git, npm, python, curl, etc.)\n- `Read` — read any file\n- `Write` — write any file\n- `Edit` — edit files with diffs\n- `Glob` — find files by pattern\n- `Grep` — search file contents\n\nClaude reads CLAUDE.md on the volume, checks the inbox, decides what to do,\ndoes it, and writes results to the outbox. The orchestrator just reads the output.\n\n### 4. Inbox/Outbox File Convention\nCommunication between Val Town and the sandbox is via files:\n- **Orchestrator → Sandbox**: writes task files to `.mail/inbox/`\n- **Sandbox → Orchestrator**: writes response files to `.mail/outbox/`\n- **Schedule requests**: agent writes `.mail/outbox/schedule.json`\n- **File attachments**: agent writes `.mail/outbox/response-{id}.files.json`\n\n## Database\n\nUses **project-scoped SQLite** (`std/sqlite/main.ts`). Do NOT change to `std/sqlite` or `std/sqlite/global.ts`.\n\n### Tables\n\n#### users\n```sql\nusers (\n id, email, name, status, volume_slug, volume_id,\n volume_region, sandbox_snapshot, created_at, updated_at\n)\n-- status: 'pending_verification' | 'verified'\n-- Existing users default to 'verified'. New users start as 'pending_verification'.\n```\n\n#### pending_emails\n```sql\npending_emails (\n id, user_id, subject, body, attachments_json, created_at\n)\n-- Stores emails received before the user has verified.\n-- Deleted after all pending emails are processed post-verification.\n```\n\n#### email_verification_tokens\n```sql\nemail_verification_tokens (\n id, user_id, token, expires_at, used_at, created_at\n)\n-- 24-hour expiry. Separate from web login magic_tokens.\n```\n\n#### magic_tokens (web login)\n```sql\nmagic_tokens (\n id, user_id, token, expires_at, used_at, created_at\n)\n-- 15-minute expiry. For web dashboard login only.\n```\n\n#### sessions\n```sql\nsessions (\n id, user_id, token, expires_at, created_at\n)\n-- 30-day web sessions. HttpOnly cookie.\n```\n\n#### scheduled_tasks\n```sql\nscheduled_tasks (\n id, user_id, title, description, next_run_at,\n schedule_description, is_recurring, status, created_at, updated_at\n)\n```\n\n#### sandbox_executions\n```sql\nsandbox_executions (\n id, user_id, trigger_type, trigger_id,\n status, turns, cost_usd, elapsed_ms, error_message,\n created_at, completed_at\n)\n```\n\n## Sandbox Lifecycle\n\n### Email Flow\n1. Email arrives → `email-handler-v2.ts`\n2. Get/create user from sender email\n3. **If user is `pending_verification`** (new or unverified):\n a. Store email in `pending_emails` table (subject, body, serialized attachments)\n b. Send verification email with magic link + welcome instructions (new users) or reminder (repeat)\n c. Stop — no sandbox execution until verified\n4. **If user is `verified`** → normal processing:\n a. Call `executeSandboxTask()`:\n - Get/create user's volume (provisions if first time)\n - Write email content to `.mail/inbox/task-{id}-{slug}.md`\n - Boot sandbox, mount user's volume\n - Run `node /opt/agent/agent-loop.mjs`\n - Agent reads CLAUDE.md, reads inbox, processes task\n - Agent writes response to `.mail/outbox/response-{id}.md`\n - Agent optionally writes `schedule.json`, `files.json`\n - Orchestrator reads outbox, shuts down sandbox\n b. Process schedule requests → insert into `scheduled_tasks`\n c. Send response email to user\n\n### Email Verification Flow\n1. New user sends first email → stored in `pending_emails`, verification email sent\n2. User clicks magic link → `GET /verify-email?token=...` in `main.ts`\n3. Token validated → user status set to `verified`\n4. All pending emails replayed through sandbox (background, non-blocking)\n5. Replies sent for each pending email\n\n### Scheduled Task Flow\n1. `task-runner-v2.ts` fires every 15 min\n2. Query `scheduled_tasks` for due tasks\n3. For each: `executeSandboxTask()` (same as above)\n4. Email results to user\n5. Reschedule recurring tasks\n\n### Volume Initialization\nFirst time a user needs a sandbox:\n1. Create a persistent volume via `Client.volumes.create()` (slug: `email-agent-user-{userId}`, region: `ord`, capacity: `1GB`)\n2. Record volume_slug in users table\n3. Boot sandbox with volume mounted at `/home/app` via `Sandbox.create({ volumes: { \"/home/app\": slug } })`\n4. Create directory structure: `.mail/inbox`, `.mail/outbox`, `memories/`, etc.\n5. Write initial CLAUDE.md, agents.md, TODO.md\n6. Install agent-loop.mjs and dependencies\n7. All subsequent runs mount the same volume — memories, files, and agent state persist\n\n## Environment Variables\n\n- **ANTHROPIC_API_KEY** — required. Passed into sandboxes for Claude API calls.\n\n## Dependencies\n\n- `jsr:@deno/sandbox` — Deno Sandbox SDK (Client for volumes, Sandbox for execution)\n- `npm:@anthropic-ai/sdk@0.39.0` — Claude API client (for scheduler.ts)\n- `https://esm.town/v/std/sqlite/main.ts` — project-scoped SQLite\n- `https://esm.town/v/std/email` — Val Town email sending\n\n## Conventions\n\n### Code Style\n- TypeScript everywhere (.ts files only)\n- File and val names in **kebab-case**\n- Always use parameterized SQL queries (`args`)\n- Emoji prefixes in console.log: 📧 📝 🧠 📊 💾 💡 ⏰ 🔄 ⚡ ✅ ❌ 📤 🔧 📋 🚀\n\n### Error Handling\n- Every sandbox execution is wrapped in try/catch\n- Task failures still send a notification email\n- Recurring tasks get rescheduled even on failure\n- Sandbox boot failures return a user-friendly error message\n\n### Sandbox Conventions\n- User volumes are named `email-agent-user-{userId}`\n- Sandbox region: `us-east-1` (configurable in orchestrator)\n- Sandbox timeout: 5 min per execution\n- Agent cycle timeout: 3 min (inside the sandbox)\n- Agent max turns: 30\n\n## Known Limitations & Future Work\n\n- **Sandbox boot latency**: ~2-5 seconds per execution. Not ideal for trivial questions. Consider a fast-path that skips the sandbox for simple queries.\n- **Concurrency**: Deno Sandbox has a 5 concurrent limit during pre-release. With multiple users emailing simultaneously, may need queuing.\n- **Volume persistence model**: Each user gets a dedicated Deno Sandbox volume (slug: `email-agent-user-{userId}`) mounted at `/home/app`. Files persist across sandbox sessions. Volumes are created via `Client.volumes.create()` and mounted via `Sandbox.create({ volumes: { \"/home/app\": slug } })`.\n- **File attachments**: Outbound file attachment support (reading files from sandbox, converting to email attachments) is stubbed but not yet implemented.\n- **No auth on HTTP admin**: The endpoints are open.\n- **No rate limiting**: No protection against flooding.\n- **Cost tracking**: Sandbox compute time + Claude API calls need monitoring. The `sandbox_executions` table logs this but there's no alerting or per-user limits yet.\n- **SSH key management**: The infrastructure is there (`.ssh/` dir on volume, SSH config in snapshot) but no user-facing flow for generating/managing keys yet.\n- **Snapshot management**: Need a process for updating the toolchain (new Node.js version, new tools, updated agent-loop.mjs) and rolling out to existing users.\n",[175],"edffb3bc-180f-11f1-a610-42dde27851f2","lib","edffb48e-180f-11f1-a610-42dde27851f2",3,["D",1772659212112],0,"directory",[182],"ede42160-180f-11f1-9915-42dde27851f2","db","ede42228-180f-11f1-9915-42dde27851f2",["D",1772659211984],[190],"5e852464-18bf-11f1-867c-42dde27851f2","admin","019cbf36-c21e-7599-8efe-0c7acaaf99fc",["D",1772734562850],[195],"pulls","forks","envVars",{"_205":59,"_207":186},"connections",{"_205":186,"_206":186},"count","unfilledCount","emptyCount",{"_21":194,"_13":195,"_133":196,"_135":-5,"_136":33,"_137":35,"_139":118,"_23":321,"_30":186,"_141":187,"_143":-5,"_145":46,"_146":-5,"_226":322,"_99":323},{"_21":314,"_13":315,"_133":316,"_135":196,"_136":33,"_137":35,"_139":118,"_23":317,"_30":186,"_141":246,"_143":318,"_145":46,"_146":-5,"_226":319,"_99":320},{"_21":306,"_13":307,"_133":308,"_135":196,"_136":33,"_137":309,"_139":118,"_23":310,"_30":171,"_141":246,"_143":311,"_145":46,"_146":-5,"_226":312,"_99":313},{"_21":189,"_13":190,"_133":191,"_135":-5,"_136":33,"_137":59,"_139":118,"_23":303,"_30":186,"_141":187,"_143":-5,"_145":46,"_146":-5,"_226":304,"_99":305},{"_21":294,"_13":295,"_133":296,"_135":191,"_136":33,"_137":297,"_139":118,"_23":298,"_30":299,"_141":246,"_143":300,"_145":46,"_146":-5,"_226":301,"_99":302},{"_21":181,"_13":182,"_133":183,"_135":-5,"_136":33,"_137":184,"_139":118,"_23":291,"_30":186,"_141":187,"_143":-5,"_145":46,"_146":-5,"_226":292,"_99":293},{"_21":282,"_13":283,"_133":284,"_135":183,"_136":33,"_137":285,"_139":118,"_23":286,"_30":287,"_141":246,"_143":288,"_145":46,"_146":-5,"_226":289,"_99":290},{"_21":274,"_13":275,"_133":276,"_135":183,"_136":33,"_137":277,"_139":118,"_23":278,"_30":255,"_141":246,"_143":279,"_145":46,"_146":-5,"_226":280,"_99":281},{"_21":266,"_13":267,"_133":268,"_135":183,"_136":33,"_137":269,"_139":118,"_23":270,"_30":59,"_141":246,"_143":271,"_145":46,"_146":-5,"_226":272,"_99":273},{"_21":259,"_13":260,"_133":261,"_135":183,"_136":33,"_137":35,"_139":118,"_23":262,"_30":186,"_141":246,"_143":263,"_145":46,"_146":-5,"_226":264,"_99":265},{"_21":250,"_13":251,"_133":252,"_135":183,"_136":33,"_137":253,"_139":118,"_23":254,"_30":255,"_141":246,"_143":256,"_145":46,"_146":-5,"_226":257,"_99":258},{"_21":241,"_13":242,"_133":243,"_135":183,"_136":33,"_137":244,"_139":118,"_23":245,"_30":59,"_141":246,"_143":247,"_145":46,"_146":-5,"_226":248,"_99":249},{"_21":174,"_13":175,"_133":176,"_135":-5,"_136":33,"_137":177,"_139":118,"_23":238,"_30":152,"_141":111,"_143":179,"_145":46,"_146":-5,"_226":239,"_99":240},{"_21":166,"_13":167,"_133":168,"_135":-5,"_136":33,"_137":169,"_139":118,"_23":235,"_30":171,"_141":111,"_143":172,"_145":46,"_146":-5,"_226":236,"_99":237},{"_21":157,"_13":158,"_133":159,"_135":-5,"_136":33,"_137":160,"_139":118,"_23":232,"_30":162,"_141":163,"_143":164,"_145":46,"_146":155,"_226":233,"_99":234},{"_21":148,"_13":149,"_133":150,"_135":-5,"_136":33,"_137":42,"_139":118,"_23":229,"_30":152,"_141":153,"_143":154,"_145":46,"_146":155,"_226":230,"_99":231},{"_21":131,"_13":132,"_133":134,"_135":-5,"_136":33,"_137":138,"_139":118,"_23":225,"_30":59,"_141":142,"_143":144,"_145":46,"_146":-5,"_226":227,"_99":228},["D",1772836608118],"children",[],[132],["D",1772840104076],[],[149],["D",1772838669058],[],[158],["D",1772837401848],[],[167],["D",1772838768498],[],[175],"dd6bcbb0-18c5-11f1-90c8-42dde27851f2","scheduler.ts","1d2551ec-1810-11f1-a610-42dde27851f2",60,["D",1772737352748],"script","/**\n * Schedule parser — converts natural language schedules into concrete timestamps.\n *\n * Two-tier approach:\n * 1. Try simple pattern matching first (fast, no API call, deterministic)\n * 2. Fall back to Claude for complex expressions\n *\n * All timestamps are returned in SQLite-compatible UTC format: \"YYYY-MM-DD HH:MM:SS\"\n * This avoids the ISO 8601 T/Z format mismatch with SQLite's datetime() function.\n */\n\nimport Anthropic from \"npm:@anthropic-ai/sdk@0.39.0\";\n\n/** Format a Date as \"YYYY-MM-DD HH:MM:SS\" (SQLite-compatible UTC) */\nfunction toSqliteUtc(date: Date): string {\n return date.toISOString().replace(\"T\", \" \").replace(/\\.\\d{3}Z$/, \"\");\n}\n\n/**\n * Try to parse simple relative time expressions without an LLM call.\n * Returns null if the expression is too complex.\n */\nfunction trySimpleParse(\n text: string,\n): { nextRunAt: string; isRecurring: boolean; interpreted: string } | null {\n const lower = text.toLowerCase().trim();\n const now = Date.now();\n\n // \"in X minutes/hours/days\"\n const relativeMatch = lower.match(\n /^in\\s+(\\d+)\\s*(minutes?|mins?|hours?|hrs?|days?)$/,\n );\n if (relativeMatch) {\n const amount = parseInt(relativeMatch[1]);\n const unit = relativeMatch[2];\n let ms = 0;\n if (unit.startsWith(\"min\")) ms = amount * 60 * 1000;\n else if (unit.startsWith(\"h\")) ms = amount * 60 * 60 * 1000;\n else if (unit.startsWith(\"d\")) ms = amount * 24 * 60 * 60 * 1000;\n\n const target = new Date(now + ms);\n return {\n nextRunAt: toSqliteUtc(target),\n isRecurring: false,\n interpreted: `In ${amount} ${unit} (${toSqliteUtc(target)} UTC)`,\n };\n }\n\n // \"in half an hour\" / \"in an hour\"\n if (\n lower === \"in half an hour\" || lower === \"in 30 minutes\" ||\n lower === \"in 30 mins\"\n ) {\n const target = new Date(now + 30 * 60 * 1000);\n return {\n nextRunAt: toSqliteUtc(target),\n isRecurring: false,\n interpreted: `In 30 minutes (${toSqliteUtc(target)} UTC)`,\n };\n }\n if (lower === \"in an hour\" || lower === \"in 1 hour\") {\n const target = new Date(now + 60 * 60 * 1000);\n return {\n nextRunAt: toSqliteUtc(target),\n isRecurring: false,\n interpreted: `In 1 hour (${toSqliteUtc(target)} UTC)`,\n };\n }\n\n // \"tomorrow\" / \"tomorrow at HH:MM\"\n const tomorrowMatch = lower.match(\n /^tomorrow(?:\\s+at\\s+(\\d{1,2})(?::(\\d{2}))?\\s*(am|pm)?)?$/,\n );\n if (tomorrowMatch) {\n const target = new Date(now);\n target.setUTCDate(target.getUTCDate() + 1);\n let hours = tomorrowMatch[1] ? parseInt(tomorrowMatch[1]) : 9;\n const minutes = tomorrowMatch[2] ? parseInt(tomorrowMatch[2]) : 0;\n const ampm = tomorrowMatch[3];\n if (ampm === \"pm\" && hours < 12) hours += 12;\n if (ampm === \"am\" && hours === 12) hours = 0;\n target.setUTCHours(hours, minutes, 0, 0);\n return {\n nextRunAt: toSqliteUtc(target),\n isRecurring: false,\n interpreted: `Tomorrow at ${hours.toString().padStart(2, \"0\")}:${\n minutes.toString().padStart(2, \"0\")\n } UTC`,\n };\n }\n\n return null;\n}\n\nconst SCHEDULE_PROMPT =\n `You are a scheduling assistant. Given a natural language time expression and the current UTC datetime, return the next occurrence.\n\nRules:\n- If no time is specified, default to 09:00 UTC\n- For recurring schedules (\"every Monday\"), return the NEXT occurrence\n- For relative times (\"in 2 hours\"), calculate from now\n- For specific dates (\"next Thursday\"), calculate the date\n- Always return UTC times\n- Format the datetime as \"YYYY-MM-DD HH:MM:SS\" (NO T separator, NO Z suffix). Example: \"2026-03-05 14:30:00\"\n\nRespond with ONLY a JSON object:\n{\n \"next_run_at\": \"2026-01-15 09:00:00\",\n \"is_recurring\": true/false,\n \"interpreted_schedule\": \"Every Monday at 09:00 UTC\"\n}`;\n\nexport async function parseSchedule(\n scheduleText: string,\n): Promise<{ nextRunAt: string; isRecurring: boolean; interpreted: string }> {\n // Try simple parsing first — fast, deterministic, no API cost\n const simple = trySimpleParse(scheduleText);\n if (simple) {\n console.log(\n `⏰ Schedule parsed locally: \"${scheduleText}\" → ${simple.nextRunAt}`,\n );\n return simple;\n }\n\n // Fall back to Claude for complex expressions\n console.log(`⏰ Schedule requires LLM: \"${scheduleText}\"`);\n const client = new Anthropic();\n const now = toSqliteUtc(new Date());\n\n const response = await client.messages.create({\n model: \"claude-sonnet-4-6\",\n max_tokens: 256,\n system: SCHEDULE_PROMPT,\n messages: [\n {\n role: \"user\",\n content:\n `Current UTC time: ${now}\\nSchedule expression: \"${scheduleText}\"`,\n },\n ],\n });\n\n const text = response.content[0].type === \"text\"\n ? response.content[0].text\n : \"\";\n const jsonMatch = text.match(/\\{[\\s\\S]*\\}/);\n\n if (!jsonMatch) {\n const fallback = toSqliteUtc(new Date(Date.now() + 60 * 60 * 1000));\n console.warn(`⏰ Schedule parse failed, falling back to +1hr: ${fallback}`);\n return {\n nextRunAt: fallback,\n isRecurring: false,\n interpreted: scheduleText,\n };\n }\n\n try {\n const parsed = JSON.parse(jsonMatch[0]);\n // Normalize the timestamp to SQLite format in case the LLM returns ISO\n let nextRunAt: string = parsed.next_run_at;\n nextRunAt = nextRunAt.replace(\"T\", \" \").replace(/Z$/, \"\").replace(\n /\\.\\d{3}$/,\n \"\",\n );\n\n return {\n nextRunAt,\n isRecurring: parsed.is_recurring ?? false,\n interpreted: parsed.interpreted_schedule ?? scheduleText,\n };\n } catch {\n const fallback = toSqliteUtc(new Date(Date.now() + 60 * 60 * 1000));\n console.warn(\n `⏰ Schedule JSON parse failed, falling back to +1hr: ${fallback}`,\n );\n return {\n nextRunAt: fallback,\n isRecurring: false,\n interpreted: scheduleText,\n };\n }\n}\n\n/**\n * Given a recurring task's schedule description, calculate the NEXT run time\n * after the current run. Used after a recurring task executes.\n */\nexport async function getNextRecurrence(\n scheduleDescription: string,\n): Promise {\n const result = await parseSchedule(scheduleDescription);\n return result.nextRunAt;\n}",[],[182,242],"e0290308-19ac-11f1-8830-42dde27851f2","sandbox-orchestrator.ts","5e84aab6-18bf-11f1-867c-42dde27851f2",71,["D",1772836571088],5,"/**\n * Sandbox Orchestrator\n *\n * Manages the lifecycle of per-user Deno Sandbox sessions:\n * 1. Ensures the user has a persistent volume (creates one if first time)\n * 2. Tries to RECONNECT to the user's existing running sandbox\n * 3. Only creates a new sandbox if none is running\n * 4. Writes task to the user's inbox on the volume\n * 5. Runs agent-loop.mjs (Claude Agent SDK with full Linux tools)\n * 6. Reads results from the outbox\n * 7. Disconnects (close) without killing — sandbox stays warm for next request\n *\n * IMPORTANT: Sandboxes use `lifetime: \"30m\"` so they persist between email\n * requests. We label each sandbox with `user-id` so we can find it later\n * via `Sandbox.list({ labels })`. The sandbox ID is also stored in the\n * users table for direct reconnection.\n *\n * This prevents the \"25 orphaned sandboxes\" problem — each user should\n * have at most ONE sandbox running at any time.\n */\n\nimport { Client, Sandbox } from \"jsr:@deno/sandbox\";\nimport {\n completeSandboxExecution,\n logSandboxExecution,\n updateUserSandboxId,\n updateUserVolume,\n type UserV2,\n} from \"../db/schema-v2.ts\";\nimport {\n buildAgentPersona,\n buildSandboxClaudeMd,\n} from \"./sandbox-claude-md.ts\";\nimport {\n AGENT_LOOP_PACKAGE_JSON,\n AGENT_LOOP_SOURCE,\n} from \"./sandbox-agent-loop.ts\";\n\n// ============ CONFIGURATION ============\n\nconst USER_HOME = \"/home/app\";\nconst AGENT_DIR = \"/home/app/agent\";\nconst VOLUME_REGION = \"ord\";\nconst VOLUME_CAPACITY = \"1GB\";\n\n// How long a sandbox stays alive after we disconnect.\n// This allows the next email from the same user to reuse the warm sandbox.\nconst SANDBOX_LIFETIME = \"30m\";\n\n// Directories to ensure exist on the user's workspace\nconst VOLUME_DIRS = [\n \".mail/inbox\",\n \".mail/outbox\",\n \".agent-results\",\n \"memories\",\n \"repos\",\n \"workspace\",\n \"ideas\",\n \"people\",\n \".ssh\",\n];\n\n// ============ TYPES ============\n\nexport interface SandboxTaskInput {\n title: string;\n description: string;\n context?: string; // serialized memories, recent interactions, etc.\n}\n\nexport interface FileAttachment {\n filename: string; // just the basename, e.g. \"report.csv\"\n path: string; // absolute path on the sandbox volume\n content: string; // base64-encoded file content\n mimeType: string; // detected MIME type\n size: number; // size in bytes\n}\n\nexport interface SandboxTaskResult {\n success: boolean;\n response: string; // agent's final output to send back to user\n filesCreated: string[]; // list of file paths the agent created/modified\n fileAttachments: FileAttachment[]; // actual file contents ready for email\n schedules: ScheduleRequest[]; // future tasks the agent wants scheduled\n exitCode: number;\n turns: number;\n costUsd: number | null;\n elapsedMs: number;\n error?: string;\n}\n\nexport interface ScheduleRequest {\n title: string;\n description: string;\n schedule: string; // natural language: \"tomorrow at 9am\", \"every Monday at 8am\"\n is_recurring: boolean;\n}\n\n// ============ VOLUME MANAGEMENT ============\n\n/**\n * Ensure the user has a persistent volume.\n * If the user already has a volume_slug in the DB, verify it exists.\n * If not, create a new volume and record it in the DB.\n *\n * Returns the volume slug to mount.\n */\nasync function ensureUserVolume(client: Client, user: UserV2): Promise {\n const volumeSlug = `email-agent-user-${user.id}`;\n\n if (user.volume_slug) {\n // User already has a volume recorded — verify it still exists\n try {\n await client.volumes.get(user.volume_slug);\n console.log(`📦 Reusing existing volume: ${user.volume_slug}`);\n return user.volume_slug;\n } catch {\n // Volume was deleted or doesn't exist — fall through to create a new one\n console.log(`⚠️ Volume ${user.volume_slug} not found, will recreate`);\n }\n }\n\n // Create a new volume for this user\n try {\n const volume = await client.volumes.create({\n slug: volumeSlug,\n region: VOLUME_REGION,\n capacity: VOLUME_CAPACITY,\n });\n console.log(\n `📦 Created new volume: ${volume.slug} (${volume.capacity} bytes)`,\n );\n } catch (err) {\n // Volume slug might already exist (e.g. DB was reset but volume persists)\n const errMsg = err instanceof Error ? err.message : String(err);\n if (errMsg.includes(\"already exists\") || errMsg.includes(\"conflict\")) {\n console.log(`📦 Volume ${volumeSlug} already exists, reusing`);\n } else {\n throw err;\n }\n }\n\n // Record the volume in the DB\n await updateUserVolume(user.id, volumeSlug, volumeSlug, VOLUME_REGION);\n\n return volumeSlug;\n}\n\n// ============ SANDBOX LIFECYCLE ============\n\n/**\n * Get or create a sandbox for the given user.\n *\n * Strategy:\n * 1. If the user has a sandbox_id in the DB, try to connect to it directly\n * 2. If that fails, list sandboxes with the user's label to find any running one\n * 3. If no running sandbox found, create a new one with lifetime + labels\n * 4. Store the sandbox ID in the DB for future reconnection\n *\n * Returns a connected Sandbox instance. Caller MUST call sandbox.close()\n * (not sandbox.kill()) when done to keep the sandbox alive for reuse.\n */\nasync function getOrCreateSandbox(\n user: UserV2,\n volumeSlug: string,\n): Promise<{ sandbox: Sandbox; isNew: boolean }> {\n const labels = { \"user-id\": String(user.id) };\n\n // Strategy 1: Try direct reconnect via stored sandbox_id\n if (user.sandbox_id) {\n try {\n console.log(\n `🔌 Attempting to reconnect to sandbox ${user.sandbox_id} for user ${user.id}...`,\n );\n const sandbox = await Sandbox.connect({ id: user.sandbox_id });\n console.log(\n `♻️ Reconnected to existing sandbox ${sandbox.id} for user ${user.id}`,\n );\n\n // Extend the lifetime so it doesn't expire while we're working\n try {\n await sandbox.extendTimeout(\"30m\");\n } catch {\n // Extension may fail if near max — that's okay\n console.log(`⏱️ Could not extend sandbox timeout (may be near max)`);\n }\n\n return { sandbox, isNew: false };\n } catch (err) {\n const errMsg = err instanceof Error ? err.message : String(err);\n console.log(\n `⚠️ Could not reconnect to sandbox ${user.sandbox_id}: ${errMsg}`,\n );\n // Clear stale sandbox_id\n await updateUserSandboxId(user.id, null);\n }\n }\n\n // Strategy 2: List sandboxes by label to find any running one for this user\n try {\n const existing = await Sandbox.list({ labels });\n if (existing.length > 0) {\n // Connect to the first (most recent) one\n const meta = existing[0];\n console.log(\n `🔍 Found existing sandbox ${meta.id} via label lookup for user ${user.id}`,\n );\n try {\n const sandbox = await Sandbox.connect({ id: meta.id });\n console.log(\n `♻️ Reconnected to labeled sandbox ${sandbox.id} for user ${user.id}`,\n );\n\n // Store this ID for faster reconnect next time\n await updateUserSandboxId(user.id, sandbox.id);\n\n // Extend lifetime\n try {\n await sandbox.extendTimeout(\"30m\");\n } catch {\n console.log(`⏱️ Could not extend sandbox timeout (may be near max)`);\n }\n\n return { sandbox, isNew: false };\n } catch (connectErr) {\n const errMsg = connectErr instanceof Error\n ? connectErr.message\n : String(connectErr);\n console.log(\n `⚠️ Found sandbox ${meta.id} but could not connect: ${errMsg}`,\n );\n // Fall through to create new\n }\n }\n } catch (listErr) {\n const errMsg = listErr instanceof Error ? listErr.message : String(listErr);\n console.log(`⚠️ Could not list sandboxes: ${errMsg}`);\n // Fall through to create new\n }\n\n // Strategy 3: Create a new sandbox\n console.log(`🆕 Creating new sandbox for user ${user.id}...`);\n const sandbox = await Sandbox.create({\n region: VOLUME_REGION,\n lifetime: SANDBOX_LIFETIME,\n labels,\n env: {\n HOME: USER_HOME,\n ANTHROPIC_API_KEY: Deno.env.get(\"ANTHROPIC_API_KEY\") || \"\",\n AGENT_MODEL: \"claude-sonnet-4-6\",\n AGENT_MAX_TURNS: \"30\",\n },\n volumes: {\n [USER_HOME]: volumeSlug,\n },\n });\n\n console.log(\n `✨ Created new sandbox ${sandbox.id} for user ${user.id} (lifetime: ${SANDBOX_LIFETIME})`,\n );\n\n // Store the sandbox ID for future reconnection\n await updateUserSandboxId(user.id, sandbox.id);\n\n return { sandbox, isNew: true };\n}\n\n// ============ VOLUME INITIALIZATION ============\n\n/**\n * Initialize a fresh user workspace with the directory structure,\n * CLAUDE.md, agent-loop, and base files.\n */\nasync function initializeVolume(sandbox: Sandbox, user: UserV2) {\n // Create directory structure\n for (const dir of VOLUME_DIRS) {\n await sandbox.sh`mkdir -p ${USER_HOME}/${dir}`;\n }\n\n // Write CLAUDE.md\n const claudeMd = buildSandboxClaudeMd({\n userName: user.name || undefined,\n });\n await sandbox.fs.writeTextFile(`${USER_HOME}/CLAUDE.md`, claudeMd);\n\n // Write agents.md persona\n const persona = buildAgentPersona({ userName: user.name || undefined });\n await sandbox.fs.writeTextFile(`${USER_HOME}/agents.md`, persona);\n\n // Write initial TODO.md\n await sandbox.fs.writeTextFile(\n `${USER_HOME}/TODO.md`,\n \"# TODO\\n\\n## Active\\n_No tasks yet._\\n\\n## Completed\\n\\n## Someday\\n\",\n );\n\n // Set up SSH directory permissions\n await sandbox.sh`chmod 700 ${USER_HOME}/.ssh`;\n\n // Write and install agent-loop\n await sandbox.sh`mkdir -p ${AGENT_DIR}`;\n await sandbox.fs.writeTextFile(\n `${AGENT_DIR}/agent-loop.mjs`,\n AGENT_LOOP_SOURCE,\n );\n await sandbox.fs.writeTextFile(\n `${AGENT_DIR}/package.json`,\n AGENT_LOOP_PACKAGE_JSON,\n );\n await sandbox.sh`cd ${AGENT_DIR} && npm install 2>&1`;\n\n console.log(`📦 Initialized workspace for user ${user.id} (${user.email})`);\n}\n\n// ============ TASK EXECUTION ============\n\n/**\n * Execute a task in the user's sandbox.\n *\n * This is the main entry point. It:\n * 1. Ensures the user has a persistent volume\n * 2. Gets or reconnects to the user's sandbox (creates only if needed)\n * 3. Initializes the workspace if this is the first use\n * 4. Writes the task to the inbox\n * 5. Runs agent-loop.mjs via bash script (captures all output)\n * 6. Reads results from the outbox\n * 7. Disconnects (close) — sandbox stays alive for next request\n */\nexport async function executeSandboxTask(\n user: UserV2,\n task: SandboxTaskInput,\n trigger: { type: \"email\" | \"scheduled_task\" | \"admin\"; id?: string },\n): Promise {\n // Log the execution\n const executionId = await logSandboxExecution({\n userId: user.id,\n triggerType: trigger.type,\n triggerId: trigger.id,\n });\n\n const startTime = Date.now();\n let sandbox: Sandbox | null = null;\n\n try {\n // Ensure the user has a persistent volume\n const client = new Client();\n const volumeSlug = await ensureUserVolume(client, user);\n\n // Get or create the user's sandbox (reuses existing if running)\n const { sandbox: sb, isNew } = await getOrCreateSandbox(user, volumeSlug);\n sandbox = sb;\n\n // Check if this workspace needs initialization (no CLAUDE.md yet)\n let needsInit = false;\n try {\n await sandbox.sh`test -f ${USER_HOME}/CLAUDE.md`;\n } catch {\n needsInit = true;\n }\n\n if (needsInit) {\n await initializeVolume(sandbox, user);\n } else if (isNew) {\n // New sandbox but volume already has CLAUDE.md — just ensure agent-loop is up to date\n await sandbox.sh`mkdir -p ${AGENT_DIR}`;\n await sandbox.fs.writeTextFile(\n `${AGENT_DIR}/agent-loop.mjs`,\n AGENT_LOOP_SOURCE,\n );\n await sandbox.fs.writeTextFile(\n `${AGENT_DIR}/package.json`,\n AGENT_LOOP_PACKAGE_JSON,\n );\n // Only npm install if node_modules doesn't exist yet\n try {\n await sandbox.sh`test -d ${AGENT_DIR}/node_modules`;\n } catch {\n await sandbox.sh`cd ${AGENT_DIR} && npm install 2>&1`;\n }\n }\n\n // Write the task to the inbox\n const taskId = Date.now();\n const taskFilename = `task-${taskId}-${slugify(task.title)}.md`;\n const taskContent = buildTaskFile(taskId, task);\n await sandbox.fs.writeTextFile(\n `${USER_HOME}/.mail/inbox/${taskFilename}`,\n taskContent,\n );\n\n // Run the agent loop via bash script to capture all output\n console.log(`🚀 Running sandbox task for user ${user.id}: \"${task.title}\"`);\n\n await sandbox.fs.writeTextFile(\n \"/tmp/run-agent.sh\",\n `cd ${USER_HOME} && node ${AGENT_DIR}/agent-loop.mjs > /tmp/agent.log 2>&1; echo \"EXIT=\\\\$?\" >> /tmp/agent.log`,\n );\n await sandbox.sh`bash /tmp/run-agent.sh`;\n\n // Read the captured output\n let agentLog = \"\";\n let exitCode = 0;\n try {\n agentLog = await sandbox.fs.readTextFile(\"/tmp/agent.log\");\n // Parse exit code from last line\n const exitMatch = agentLog.match(/EXIT=(\\d+)/);\n exitCode = exitMatch ? parseInt(exitMatch[1], 10) : 0;\n console.log(`Sandbox output:\\n${agentLog}`);\n } catch {\n console.log(\"Could not read agent log\");\n }\n\n // Read results from outbox\n const response = await readOutboxResponse(sandbox, taskId);\n const filesCreated = await readOutboxFiles(sandbox, taskId);\n const schedules = await readOutboxSchedules(sandbox);\n const cycleResult = await readLatestCycleResult(sandbox);\n\n // Read actual file contents for email attachments (before closing sandbox!)\n let fileAttachments: FileAttachment[] = [];\n if (filesCreated.length > 0) {\n console.log(`📎 Reading ${filesCreated.length} file(s) for email attachment...`);\n fileAttachments = await readFileAttachments(sandbox, filesCreated);\n console.log(`📎 Successfully read ${fileAttachments.length} attachment(s)`);\n }\n\n const elapsedMs = Date.now() - startTime;\n const turns = cycleResult?.turns || 0;\n const costUsd = cycleResult?.cost_usd ?? null;\n\n // Disconnect but keep sandbox alive for next request\n await sandbox.close();\n sandbox = null; // Prevent double-close in finally\n\n // Log completion\n await completeSandboxExecution(executionId, {\n status: exitCode === 0 ? \"success\" : \"failure\",\n turns,\n costUsd: costUsd ?? undefined,\n elapsedMs,\n errorMessage: exitCode !== 0 ? agentLog : undefined,\n });\n\n return {\n success: exitCode === 0,\n response: response || `Task \"${task.title}\" was processed.`,\n filesCreated,\n fileAttachments,\n schedules,\n exitCode,\n turns,\n costUsd,\n elapsedMs,\n };\n } catch (error) {\n const elapsedMs = Date.now() - startTime;\n const errMsg = error instanceof Error ? error.message : String(error);\n\n // Try to close cleanly (not kill) so sandbox can be reused\n if (sandbox) {\n try {\n await sandbox.close();\n } catch {\n // Ignore close errors\n }\n }\n\n await completeSandboxExecution(executionId, {\n status: \"failure\",\n elapsedMs,\n errorMessage: errMsg,\n });\n\n return {\n success: false,\n response:\n `I encountered an error while processing your request: ${errMsg}`,\n filesCreated: [],\n fileAttachments: [],\n schedules: [],\n exitCode: 1,\n turns: 0,\n costUsd: null,\n elapsedMs,\n error: errMsg,\n };\n }\n}\n\n// ============ READING RESULTS ============\n\n/**\n * Read the agent's response from the outbox.\n */\nasync function readOutboxResponse(\n sandbox: Sandbox,\n taskId: number,\n): Promise {\n const responsePath = `${USER_HOME}/.mail/outbox/response-${taskId}.md`;\n try {\n return await sandbox.fs.readTextFile(responsePath);\n } catch {\n // Agent might not have written to the exact expected path.\n // Try to find any markdown response file in the outbox.\n try {\n const fileList = await sandbox\n .sh`ls -t ${USER_HOME}/.mail/outbox/*.md 2>/dev/null || true`.text();\n const files = fileList.trim().split(\"\\n\").filter(Boolean);\n if (files.length > 0) {\n return await sandbox.fs.readTextFile(files[0]);\n }\n } catch {\n // Ignore\n }\n return null;\n }\n}\n\n/**\n * Read list of files the agent created for the user.\n */\nasync function readOutboxFiles(\n sandbox: Sandbox,\n taskId: number,\n): Promise {\n try {\n const json = await sandbox.fs.readTextFile(\n `${USER_HOME}/.mail/outbox/response-${taskId}.files.json`,\n );\n return JSON.parse(json);\n } catch {\n return [];\n }\n}\n\n/**\n * Read schedule requests from the agent.\n */\nasync function readOutboxSchedules(\n sandbox: Sandbox,\n): Promise {\n try {\n const json = await sandbox.fs.readTextFile(\n `${USER_HOME}/.mail/outbox/schedule.json`,\n );\n const schedules = JSON.parse(json);\n\n // Clean up the schedule file after reading\n await sandbox.sh`rm -f ${USER_HOME}/.mail/outbox/schedule.json`;\n\n return Array.isArray(schedules) ? schedules : [];\n } catch {\n return [];\n }\n}\n\n/**\n * Read the latest cycle result from .agent-results/\n */\nasync function readLatestCycleResult(\n sandbox: Sandbox,\n): Promise<{ turns: number; cost_usd: number | null; status: string } | null> {\n try {\n const latestFile = await sandbox\n .sh`ls -t ${USER_HOME}/.agent-results/ | head -1`.text();\n const filename = latestFile.trim();\n if (!filename) return null;\n\n const json = await sandbox.fs.readTextFile(\n `${USER_HOME}/.agent-results/${filename}`,\n );\n return JSON.parse(json);\n } catch {\n return null;\n }\n}\n\n// ============ FILE ATTACHMENT READING ============\n\n/**\n * Read the actual file contents from the sandbox for email attachment.\n * Reads each file path, base64-encodes it, and detects the MIME type.\n * Skips files that can't be read (e.g. deleted, too large, etc).\n */\nasync function readFileAttachments(\n sandbox: Sandbox,\n filePaths: string[],\n): Promise {\n const attachments: FileAttachment[] = [];\n const MAX_ATTACHMENT_SIZE = 10 * 1024 * 1024; // 10MB per file\n\n for (const filePath of filePaths) {\n try {\n // Get file size first to avoid reading huge files\n const statOutput = await sandbox\n .sh`stat -c '%s' ${filePath} 2>/dev/null`.text();\n const fileSize = parseInt(statOutput.trim(), 10);\n\n if (isNaN(fileSize) || fileSize === 0) {\n console.log(`⚠️ Skipping empty/unreadable file: ${filePath}`);\n continue;\n }\n\n if (fileSize > MAX_ATTACHMENT_SIZE) {\n console.log(\n `⚠️ Skipping oversized file (${(fileSize / 1024 / 1024).toFixed(1)}MB): ${filePath}`,\n );\n continue;\n }\n\n // Read the file as base64 via the sandbox shell\n const base64Content = await sandbox\n .sh`base64 -w 0 ${filePath} 2>/dev/null`.text();\n\n if (!base64Content.trim()) {\n console.log(`⚠️ Could not base64-encode file: ${filePath}`);\n continue;\n }\n\n // Extract filename from path\n const filename = filePath.split(\"/\").pop() || \"attachment\";\n\n // Detect MIME type\n const mimeType = guessMimeType(filename);\n\n attachments.push({\n filename,\n path: filePath,\n content: base64Content.trim(),\n mimeType,\n size: fileSize,\n });\n\n console.log(\n `📎 Read attachment: ${filename} (${mimeType}, ${(fileSize / 1024).toFixed(1)}KB)`,\n );\n } catch (err) {\n const errMsg = err instanceof Error ? err.message : String(err);\n console.log(`⚠️ Could not read file ${filePath}: ${errMsg}`);\n }\n }\n\n return attachments;\n}\n\n/**\n * Guess the MIME type from a filename extension.\n */\nfunction guessMimeType(filename: string): string {\n const ext = filename.split(\".\").pop()?.toLowerCase() || \"\";\n const mimeMap: Record = {\n // Text\n txt: \"text/plain\",\n md: \"text/markdown\",\n csv: \"text/csv\",\n html: \"text/html\",\n htm: \"text/html\",\n css: \"text/css\",\n xml: \"text/xml\",\n // Code\n js: \"text/javascript\",\n mjs: \"text/javascript\",\n ts: \"text/typescript\",\n json: \"application/json\",\n yaml: \"text/yaml\",\n yml: \"text/yaml\",\n toml: \"text/plain\",\n // Documents\n pdf: \"application/pdf\",\n doc: \"application/msword\",\n docx: \"application/vnd.openxmlformats-officedocument.wordprocessingml.document\",\n xls: \"application/vnd.ms-excel\",\n xlsx: \"application/vnd.openxmlformats-officedocument.spreadsheetml.sheet\",\n ppt: \"application/vnd.ms-powerpoint\",\n pptx: \"application/vnd.openxmlformats-officedocument.presentationml.presentation\",\n // Images\n png: \"image/png\",\n jpg: \"image/jpeg\",\n jpeg: \"image/jpeg\",\n gif: \"image/gif\",\n svg: \"image/svg+xml\",\n webp: \"image/webp\",\n // Archives\n zip: \"application/zip\",\n tar: \"application/x-tar\",\n gz: \"application/gzip\",\n // Other\n ics: \"text/calendar\",\n vcf: \"text/vcard\",\n };\n return mimeMap[ext] || \"application/octet-stream\";\n}\n\n// ============ HELPERS ============\n\n/**\n * Build the task markdown file that goes into the inbox.\n */\nfunction buildTaskFile(taskId: number, task: SandboxTaskInput): string {\n const parts: string[] = [\n `# ${task.title}`,\n \"\",\n task.description,\n \"\",\n ];\n\n if (task.context) {\n parts.push(\"## Additional Context\", \"\", task.context, \"\");\n }\n\n parts.push(\n \"## Instructions\",\n \"\",\n \"Complete this task. Write your response (what to email back to the user) to:\",\n `\\`${USER_HOME}/.mail/outbox/response-${taskId}.md\\``,\n \"\",\n \"If you create any files the user should receive as email attachments, list their absolute paths in:\",\n `\\`${USER_HOME}/.mail/outbox/response-${taskId}.files.json\\``,\n \"(JSON array of absolute paths)\",\n \"\",\n \"If you need to schedule future work, write to:\",\n `\\`${USER_HOME}/.mail/outbox/schedule.json\\``,\n \"(JSON array of objects with: title, description, schedule, is_recurring)\",\n );\n\n return parts.join(\"\\n\");\n}\n\n/**\n * Convert a string to a URL-safe slug.\n */\nfunction slugify(text: string): string {\n return text\n .toLowerCase()\n .replace(/[^a-z0-9]+/g, \"-\")\n .replace(/^-|-$/g, \"\")\n .slice(0, 50);\n}",[],[182,251],"5e8478de-18bf-11f1-867c-42dde27851f2","sandbox-claude-md.ts","5e84796a-18bf-11f1-867c-42dde27851f2",["D",1772734562850],"/**\n * Builds the CLAUDE.md file that lives on each user's persistent volume.\n *\n * This is the instruction set that the agent-loop.mjs reads on every boot.\n * It tells Claude how to operate in the user's workspace: where to find things,\n * how to process tasks, and how to communicate results back.\n *\n * The user's personal instructions (equivalent of the old system_prompt column)\n * are embedded in a section at the bottom. Claude can also edit this file\n * directly — if the user says \"be more casual\", Claude updates CLAUDE.md itself.\n */\n\nconst USER_HOME = \"/home/app\";\n\n/**\n * Build the full CLAUDE.md content for a user's volume.\n *\n * @param userPreferences - Optional user-specific instructions (migrated from\n * the old system_prompt, or accumulated by Claude over time). If empty,\n * Claude will build this section organically as the user gives instructions.\n */\nexport function buildSandboxClaudeMd(opts?: {\n userPreferences?: string;\n userName?: string;\n}): string {\n const prefs = opts?.userPreferences?.trim();\n const nameSection = opts?.userName\n ? `The user's name is **${opts.userName}**.`\n : \"\";\n\n return `# Agent Workspace\n\nYou are a personal AI agent. You communicate with your user via email.\nYou wake up periodically to process new messages and complete tasks.\n${nameSection}\n\n## Your Environment\n\n- **Home**: ${USER_HOME}\n- **Inbox** (new work arrives here): \\`${USER_HOME}/.mail/inbox/\\`\n- **Outbox** (write your responses here): \\`${USER_HOME}/.mail/outbox/\\`\n- **Memories & Knowledge**: \\`${USER_HOME}/memories/\\`\n- **Task List**: \\`${USER_HOME}/TODO.md\\`\n- **Ideas**: \\`${USER_HOME}/ideas/\\`\n- **People**: \\`${USER_HOME}/people/\\`\n- **Git Repos**: \\`${USER_HOME}/repos/\\`\n- **SSH Keys**: \\`${USER_HOME}/.ssh/\\`\n- **Workspace** (scratch area): \\`${USER_HOME}/workspace/\\`\n\n## Workflow\n\nWhen you wake up for a work cycle:\n\n1. **Check inbox**: Read any new files in \\`.mail/inbox/\\`. Each file is a task or message.\n2. **For each inbox item**:\n a. Read the task/message carefully\n b. Consult your \\`memories/\\` for relevant context about the user\n c. Check \\`TODO.md\\` for related pending work\n d. Do the work — run commands, edit files, fetch URLs, clone repos, whatever is needed\n e. Write your response to the path specified in the task file (usually \\`.mail/outbox/response-{id}.md\\`)\n f. If you created files the user should receive as email attachments, list them in \\`.mail/outbox/response-{id}.files.json\\` (JSON array of absolute paths)\n3. **Check TODO.md**: Process any pending items that are due or actionable now\n4. **Update your knowledge**: If you learned new facts about the user, update files in \\`memories/\\`\n5. **Schedule future work**: If something needs to happen later, write to \\`.mail/outbox/schedule.json\\`:\n \\`\\`\\`json\n [\n {\n \"title\": \"Weekly stock report\",\n \"description\": \"Analyze the user's portfolio and email a summary\",\n \"schedule\": \"every Monday at 9am\",\n \"is_recurring\": true\n }\n ]\n \\`\\`\\`\n6. **Exit** when all work is done. You will wake up again when needed.\n\n## Memory Organization\n\nStore knowledge about the user in \\`memories/\\` as markdown files:\n- One file per topic or category — you decide the organization\n- Examples: \\`memories/preferences.md\\`, \\`memories/work.md\\`, \\`memories/family.md\\`, \\`memories/projects.md\\`\n- Keep files concise and current — remove stale information\n- You can restructure/reorganize files as the knowledge grows\n- When you learn something new, update or create the appropriate file\n\n## TODO.md Format\n\nMaintain \\`TODO.md\\` as a simple task list:\n\\`\\`\\`markdown\n# TODO\n\n## Active\n- [ ] Task description here\n- [ ] Another task (due: 2026-03-10)\n\n## Completed\n- [x] Previously completed task (done: 2026-03-05)\n\n## Someday\n- [ ] Low-priority items\n\\`\\`\\`\n\nUpdate this file as you complete work or receive new tasks.\n\n## People Directory\n\nTrack people the user mentions in \\`people/\\`:\n- One file per person: \\`people/sarah-chen.md\\`\n- Include: relationship, contact info, key facts, how they connect to the user\n- Update as you learn more\n\n## Ideas\n\nStore the user's explicitly captured ideas in \\`ideas/\\`:\n- One file per idea: \\`ideas/2026-03-05-app-concept.md\\`\n- Only store ideas the user explicitly asks to save\n- Include context, related thoughts, and any follow-up actions\n\n## Git & SSH\n\n- Repos are cloned to \\`repos/\\` — they persist between sessions\n- SSH keys live in \\`.ssh/\\` — generate with \\`ssh-keygen -t ed25519\\`\n- Configure git identity: \\`git config --global user.name \"...\"\\` and \\`git config --global user.email \"...\"\\`\n- For SSH host verification: \\`ssh-keyscan github.com >> ~/.ssh/known_hosts\\`\n\n## Responding to Tasks\n\nWhen writing response files:\n- Be clear, readable, and appropriately detailed\n- Format as markdown — it will be emailed to the user\n- If you need to send files, create them and list paths in the \\`.files.json\\`\n- Be efficient — don't over-explain unless asked\n\n## Scheduling Future Work\n\nThe \\`schedule.json\\` file in your outbox is how you tell the system to wake you up later.\nEach entry needs:\n- \\`title\\`: Short name for the task\n- \\`description\\`: What to do when it fires\n- \\`schedule\\`: Natural language time — \"tomorrow at 9am\", \"every Monday at 8am\", \"in 2 hours\"\n- \\`is_recurring\\`: true for repeating tasks, false for one-time\n\n## Updating These Instructions\n\nIf the user gives you standing instructions (\"be more casual\", \"always use bullet points\",\n\"call me Paul\"), **edit this file directly** — add or update the \"User's Personal Instructions\"\nsection below. This is how preferences persist across sessions.\n\n---\n\n## User's Personal Instructions\n\n${prefs || \"_No custom instructions yet. The user can tell you their preferences and you'll save them here._\"}\n`;\n}\n\n/**\n * Build the agents.md persona file for the sandbox.\n * This is loaded by agent-loop.mjs as supplementary identity/persona info.\n */\nexport function buildAgentPersona(opts?: {\n userName?: string;\n}): string {\n return `## Persona\n\nYou are a helpful, efficient personal assistant. You communicate via email with your user.\n\n### Core traits:\n- **Thorough**: When given a task, do it completely. Don't leave loose ends.\n- **Context-aware**: Always check your memories before responding. Personalize.\n- **Proactive**: If you notice something useful while working, mention it.\n- **Honest**: If you can't do something or aren't sure, say so.\n- **Organized**: Keep your workspace tidy. Update memories, clean up TODO.md.\n\n### What you can do:\n- Search the web for current information\n- Clone and work with git repositories\n- Run code (Node.js, Python, shell scripts)\n- Read and write files\n- Install packages (npm, pip, apt)\n- Manage SSH keys\n- Create documents, reports, data exports\n- Analyze data and code\n\n### What you should NOT do:\n- Never expose SSH private keys in your responses\n- Never include API keys or secrets in response files\n- Don't modify files outside your home directory unless explicitly asked\n- Don't run destructive commands (rm -rf /) without clear user intent\n${opts?.userName ? `\\n### User\\nThe user's name is ${opts.userName}.` : \"\"}\n`;\n}\n",[],[182,260],"7f57aa00-19a9-11f1-828e-42dde27851f2","sandbox-agent-loop.ts","5e845106-18bf-11f1-867c-42dde27851f2",63,["D",1772835120174],"/**\n * Agent loop source code — runs INSIDE the Deno Sandbox.\n *\n * This file is NOT executed by Val Town directly. It exports the source code\n * as a string constant that gets written to the user's volume and executed\n * with Node.js inside the sandbox.\n *\n * Uses the Claude Agent SDK (@anthropic-ai/claude-agent-sdk) with\n * Bash/Read/Write/Edit/Glob/Grep tools in bypassPermissions mode.\n *\n * The SDK's query() returns an async generator of messages. We iterate\n * through them to completion, then read the result message for cost/turns.\n *\n * Ref: https://platform.claude.com/docs/en/agent-sdk/overview\n * Ref: https://platform.claude.com/docs/en/agent-sdk/typescript\n */\n\n/**\n * The agent-loop.mjs source code.\n *\n * Environment variables expected:\n * - ANTHROPIC_API_KEY: For Claude API calls\n * - HOME: /home/app\n * - AGENT_MODEL: Claude model to use (default: claude-sonnet-4-6)\n * - AGENT_MAX_TURNS: Max tool-use turns (default: 30)\n */\nexport const AGENT_LOOP_SOURCE = `#!/usr/bin/env node\n/**\n * agent-loop.mjs — Runs inside the Deno Sandbox\n *\n * Reads tasks from .mail/inbox/, processes them with Claude Agent SDK,\n * writes results to .mail/outbox/, then exits.\n */\n\nimport { query } from \"@anthropic-ai/claude-agent-sdk\";\nimport { readdir, readFile, writeFile, unlink, mkdir } from \"node:fs/promises\";\nimport { join } from \"node:path\";\n\nconst HOME = process.env.HOME || \"/home/app\";\nconst MODEL = process.env.AGENT_MODEL || \"claude-sonnet-4-6\";\nconst MAX_TURNS = parseInt(process.env.AGENT_MAX_TURNS || \"30\", 10);\n\nconst INBOX = join(HOME, \".mail\", \"inbox\");\nconst OUTBOX = join(HOME, \".mail\", \"outbox\");\nconst RESULTS_DIR = join(HOME, \".agent-results\");\n\n// Ensure directories exist\nasync function ensureDirs() {\n for (const dir of [INBOX, OUTBOX, RESULTS_DIR, join(HOME, \"memories\"), join(HOME, \"repos\"), join(HOME, \"workspace\"), join(HOME, \"ideas\"), join(HOME, \"people\"), join(HOME, \".ssh\")]) {\n await mkdir(dir, { recursive: true });\n }\n}\n\n// Read CLAUDE.md for system instructions\nasync function readSystemPrompt() {\n const claudeMdPath = join(HOME, \"CLAUDE.md\");\n try {\n return await readFile(claudeMdPath, \"utf-8\");\n } catch {\n return \"You are a helpful personal AI agent. Check .mail/inbox/ for tasks.\";\n }\n}\n\n// Get all pending tasks from inbox\nasync function getInboxTasks() {\n try {\n const files = await readdir(INBOX);\n const tasks = [];\n for (const file of files.sort()) {\n if (file.startsWith(\".\")) continue;\n const content = await readFile(join(INBOX, file), \"utf-8\");\n tasks.push({ filename: file, content, path: join(INBOX, file) });\n }\n return tasks;\n } catch {\n return [];\n }\n}\n\n// Read TODO.md if it exists\nasync function readTodo() {\n try {\n return await readFile(join(HOME, \"TODO.md\"), \"utf-8\");\n } catch {\n return null;\n }\n}\n\n// Write heartbeat for monitoring\nasync function writeHeartbeat(status) {\n const heartbeatPath = join(HOME, \".agent-heartbeat\");\n const data = JSON.stringify({\n status,\n timestamp: new Date().toISOString(),\n model: MODEL,\n });\n await writeFile(heartbeatPath, data, \"utf-8\");\n}\n\n// Save cycle result for the orchestrator to read\nasync function saveCycleResult(result) {\n const filename = \"cycle-\" + Date.now() + \".json\";\n await writeFile(join(RESULTS_DIR, filename), JSON.stringify(result, null, 2), \"utf-8\");\n}\n\n// ============ MAIN ============\n\nasync function main() {\n const startTime = Date.now();\n\n await ensureDirs();\n await writeHeartbeat(\"starting\");\n\n const systemPrompt = await readSystemPrompt();\n const inboxTasks = await getInboxTasks();\n const todo = await readTodo();\n\n if (inboxTasks.length === 0 && !todo) {\n console.log(\"No tasks in inbox and no TODO.md. Nothing to do.\");\n await writeHeartbeat(\"idle\");\n await saveCycleResult({\n status: \"idle\",\n turns: 0,\n cost_usd: 0,\n elapsed_ms: Date.now() - startTime,\n });\n return;\n }\n\n // Build the prompt for this cycle\n const parts = [];\n\n if (inboxTasks.length > 0) {\n parts.push(\"## New Tasks in Inbox\\\\n\");\n for (const task of inboxTasks) {\n parts.push(\"### \" + task.filename + \"\\\\n\\\\n\" + task.content + \"\\\\n\");\n }\n }\n\n if (todo) {\n parts.push(\"## Current TODO.md\\\\n\\\\n\" + todo + \"\\\\n\");\n }\n\n parts.push(\"---\");\n parts.push(\"Process the above tasks following the instructions in CLAUDE.md. When done, exit.\");\n\n const userPrompt = parts.join(\"\\\\n\\\\n\");\n\n console.log(\"Starting agent cycle: \" + inboxTasks.length + \" inbox task(s), model=\" + MODEL + \", maxTurns=\" + MAX_TURNS);\n await writeHeartbeat(\"running\");\n\n let turns = 0;\n let costUsd = null;\n let resultText = \"\";\n let errorMessage = null;\n\n try {\n // query() returns an async generator of SDKMessage\n const agentQuery = query({\n prompt: userPrompt,\n options: {\n model: MODEL,\n maxTurns: MAX_TURNS,\n allowedTools: [\"Bash\", \"Read\", \"Write\", \"Edit\", \"Glob\", \"Grep\", \"WebFetch\", \"WebSearch\"],\n permissionMode: \"bypassPermissions\",\n allowDangerouslySkipPermissions: true,\n cwd: HOME,\n systemPrompt: systemPrompt,\n },\n });\n\n // Iterate through all messages\n for await (const message of agentQuery) {\n // Log assistant messages for debugging\n if (message.type === \"assistant\" && message.message?.content) {\n for (const block of message.message.content) {\n if (\"text\" in block && block.text) {\n console.log(\"[assistant]\", block.text.substring(0, 200));\n } else if (\"name\" in block) {\n console.log(\"[tool_use]\", block.name);\n }\n }\n }\n\n // Capture the final result\n if (message.type === \"result\") {\n resultText = message.result || \"\";\n turns = message.num_turns || 0;\n costUsd = message.total_cost_usd ?? null;\n console.log(\"Agent cycle completed: \" + turns + \" turns, cost=$\" + (costUsd?.toFixed(4) ?? \"unknown\"));\n\n if (message.is_error) {\n errorMessage = \"Agent ended with error: \" + message.subtype;\n }\n }\n }\n\n // Clean up processed inbox files\n for (const task of inboxTasks) {\n try {\n await unlink(task.path);\n console.log(\"Cleaned up inbox: \" + task.filename);\n } catch (e) {\n console.warn(\"Failed to clean up \" + task.filename + \": \" + e.message);\n }\n }\n\n await writeHeartbeat(\"completed\");\n } catch (err) {\n errorMessage = err.message || String(err);\n console.error(\"Agent cycle error: \" + errorMessage);\n await writeHeartbeat(\"error\");\n }\n\n const elapsed = Date.now() - startTime;\n\n await saveCycleResult({\n status: errorMessage ? \"error\" : \"success\",\n turns,\n cost_usd: costUsd,\n elapsed_ms: elapsed,\n error: errorMessage,\n result_text: resultText,\n tasks_processed: inboxTasks.length,\n });\n\n console.log(\"Cycle finished in \" + elapsed + \"ms\");\n}\n\nmain().catch((err) => {\n console.error(\"Fatal error in agent loop:\", err);\n process.exit(1);\n});\n`;\n\n/**\n * package.json for the agent loop.\n *\n * Uses @anthropic-ai/claude-agent-sdk which includes the full agent\n * runtime with built-in tools (Bash, Read, Write, Edit, Glob, Grep).\n */\nexport const AGENT_LOOP_PACKAGE_JSON = JSON.stringify(\n {\n name: \"agent-loop\",\n version: \"1.0.0\",\n type: \"module\",\n dependencies: {\n \"@anthropic-ai/claude-agent-sdk\": \"latest\",\n },\n },\n null,\n 2,\n);",[],[182,267],"0a4b7730-19b5-11f1-909d-42dde27851f2","pages.ts","c96fce04-19b0-11f1-aefd-42dde27851f2",122,["D",1772840077752],"/**\n * HTML page templates for the web interface.\n */\n\nimport type { UserV2 } from \"../db/schema-v2.ts\";\n\nfunction baseLayout(title: string, content: string, user?: UserV2 | null): string {\n const adminLink = user?.is_admin ? `Admin` : \"\";\n const nav = user\n ? ``\n : ``;\n\n return `\n\n\n \n \n ${escapeHtml(title)} — Email Agent\n \n\n\n ${nav}\n

${content}

\n\n`;\n}\n\nfunction escapeHtml(s: string): string {\n return s\n .replace(/&/g, \"&\")\n .replace(//g, \">\")\n .replace(/\"/g, \""\");\n}\n\nfunction statusBadge(status: string): string {\n const cls =\n status === \"success\" || status === \"completed\"\n ? \"badge-success\"\n : status === \"failure\" || status === \"failed\" || status === \"timeout\"\n ? \"badge-error\"\n : status === \"pending\"\n ? \"badge-pending\"\n : status === \"running\" || status === \"started\"\n ? \"badge-running\"\n : \"badge-pending\";\n return `${escapeHtml(status)}`;\n}\n\nfunction triggerBadge(type: string): string {\n const cls = type === \"email\" ? \"badge-email\" : \"badge-scheduled\";\n return `${escapeHtml(type)}`;\n}\n\nfunction timeAgo(dateStr: string): string {\n const date = new Date(dateStr.endsWith(\"Z\") ? dateStr : dateStr + \"Z\");\n const now = Date.now();\n const diffMs = now - date.getTime();\n if (diffMs < 0) return \"in the future\";\n const mins = Math.floor(diffMs / 60000);\n if (mins < 1) return \"just now\";\n if (mins < 60) return `${mins}m ago`;\n const hours = Math.floor(mins / 60);\n if (hours < 24) return `${hours}h ago`;\n const days = Math.floor(hours / 24);\n if (days < 30) return `${days}d ago`;\n return date.toLocaleDateString();\n}\n\n// ============ PUBLIC PAGES ============\n\nexport function homePage(user: UserV2 | null): string {\n const mailtoSubject = encodeURIComponent(\"Hello!\");\n const mailtoBody = encodeURIComponent(\"Hi Email Agent! I'd like to get started. Remember that I found you from your homepage.\");\n const mailtoLink = `mailto:memory-do@valtown.email?subject=${mailtoSubject}&body=${mailtoBody}`;\n\n const ctaButton = user\n ? `Go to Dashboard →`\n : `Send Your First Email →`;\n\n const secondaryAction = user\n ? \"\"\n : `

Already have an account? Sign in

`;\n\n return baseLayout(\n \"Home\",\n `\n

📧 Email Agent

Your personal AI assistant that lives in your inbox. Send an email, get things done.

\n memory-do@valtown.email\n
\n ${ctaButton}\n ${secondaryAction}\n

\n\n

How It Works

Send an email

Write anything — a task, a memory, a question, an idea.

Agent processes it

Claude boots your personal sandbox with full Linux tools and persistent memory.

Get a reply

Results, follow-ups, and file attachments — all delivered back to your inbox.

\n\n

What It Can Do

🧠

Remember things

\"I work at Acme Corp\" · \"My friend Sarah is at Google\" · \"I like chickens\"

✅

Manage tasks

\"Remind me Thursday to call the dentist\" · \"Every Monday, summarize my tasks\"

🔍

Research & fetch

\"What's the latest on WebAssembly?\" · Fetches live data with curl, reads docs, writes reports.

💻

Run code

Full Bash, Node.js, Python, git, npm — your agent has a real dev environment.

📎

Handle files

Send attachments in, get generated files back. CSV, JSON, code, reports.

🎨

Personalize

\"Call me Paul\" · \"Be more casual\" · Your agent learns and remembers your preferences.

\n `,\n user,\n );\n}\n\n// ============ AUTH PAGES ============\n\nexport function loginPage(message?: string, error?: string): string {\n const alert = message\n ? `

${escapeHtml(message)}

`\n : error\n ? `

${escapeHtml(error)}

`\n : \"\";\n\n return baseLayout(\n \"Sign In\",\n `\n

Sign in to Email Agent

Enter the email address you use with memory-do@valtown.email and we'll send you a magic link.

\n ${alert}\n \n

\n `,\n );\n}\n\nexport function magicLinkSentPage(email: string): string {\n return baseLayout(\n \"Check Your Email\",\n `\n

📬

Check your email

We sent a sign-in link to ${escapeHtml(email)}

Click the link in the email to sign in. It expires in 15 minutes.

\n
\n Try a different email\n

\n `,\n );\n}\n\nexport function emailVerifiedPage(email: string): string {\n return baseLayout(\n \"Email Verified\",\n `\n

🎉

Email verified!

Your email address ${escapeHtml(email)} has been verified.

Your saved emails are now being processed. You'll receive replies in your inbox shortly.

\n
\n

You can now sign in to your dashboard to see your executions and scheduled tasks.

\n `,\n );\n}\n\nexport function authErrorPage(message: string): string {\n return baseLayout(\n \"Auth Error\",\n `\n

⚠️

Link expired or invalid

${escapeHtml(message)}

\n
\n Try Again\n

\n `,\n );\n}\n\n// ============ DASHBOARD ============\n\ninterface DashboardData {\n user: UserV2;\n executions: Record[];\n scheduledTasks: Record[];\n totalExecutions: number;\n totalCost: number;\n}\n\nexport function dashboardPage(data: DashboardData): string {\n const { user, executions, scheduledTasks, totalExecutions, totalCost } = data;\n\n const pendingTasks = scheduledTasks.filter(\n (t) => t.status === \"pending\",\n ).length;\n const successRate =\n totalExecutions > 0\n ? Math.round(\n (executions.filter((e) => e.status === \"success\").length /\n Math.min(executions.length, totalExecutions)) *\n 100,\n )\n : 0;\n\n // Stats cards\n const stats = `\n

${totalExecutions}

Total Executions

${pendingTasks}

Pending Tasks

$${totalCost.toFixed(2)}

Total Cost

${successRate}%

Success Rate (recent)

\n `;\n\n // Executions table\n let executionsHtml: string;\n if (executions.length === 0) {\n executionsHtml = `

No executions yet.

Send an email to memory-do@valtown.email to get started!

`;\n } else {\n const rows = executions\n .map((e) => {\n const createdAt = (e.created_at as string) || \"\";\n const elapsed = e.elapsed_ms\n ? `${((e.elapsed_ms as number) / 1000).toFixed(1)}s`\n : \"—\";\n const cost = e.cost_usd\n ? `$${(e.cost_usd as number).toFixed(4)}`\n : \"—\";\n const turns = e.turns ?? \"—\";\n const errorMsg = e.error_message\n ? `

${escapeHtml((e.error_message as string).slice(0, 80))}

`\n : \"\";\n\n return `\n ${triggerBadge(e.trigger_type as string)}\n ${statusBadge(e.status as string)}${errorMsg}\n ${turns}\n ${elapsed}\n ${cost}\n ${timeAgo(createdAt)}\n `;\n })\n .join(\"\");\n\n executionsHtml = `\n

\n \n \n \n \n ${rows}\n

Trigger	Status	Turns	Duration	Cost	When

`;\n }\n\n // Scheduled tasks table\n let tasksHtml: string;\n if (scheduledTasks.length === 0) {\n tasksHtml = `

No scheduled tasks.

`;\n } else {\n const taskRows = scheduledTasks\n .map((t) => {\n const nextRun = (t.next_run_at as string) || \"\";\n const recurring = (t.is_recurring as number) === 1 ? \"🔄 Yes\" : \"—\";\n return `\n ${escapeHtml(t.title as string)}\n ${statusBadge(t.status as string)}\n ${recurring}\n ${escapeHtml(t.schedule_description as string || \"—\")}\n ${timeAgo(nextRun)}\n `;\n })\n .join(\"\");\n\n tasksHtml = `\n

\n \n \n \n \n ${taskRows}\n

Title	Status	Recurring	Schedule	Next Run

`;\n }\n\n // Account info\n const volumeInfo = user.volume_slug\n ? `Active ${escapeHtml(user.volume_slug)}`\n : `Not provisioned`;\n\n return baseLayout(\n \"Dashboard\",\n `\n

Dashboard

Welcome back${user.name ? `, ${escapeHtml(user.name)}` : \"\"}.

\n\n ${stats}\n\n

Recent Executions

\n ${executionsHtml}\n

\n\n

Scheduled Tasks

\n ${tasksHtml}\n

\n\n

Account

\n \n \n \n \n \n

Email	${escapeHtml(user.email)}
Name	${user.name ? escapeHtml(user.name) : '—'}
Sandbox Volume	${volumeInfo}
Member since	${new Date(user.created_at.endsWith(\"Z\") ? user.created_at : user.created_at + \"Z\").toLocaleDateString()}

\n `,\n user,\n );\n}\n\n// ============ ADMIN DASHBOARD ============\n\ninterface AdminDashboardData {\n currentUser: UserV2;\n allUsers: Record[];\n recentExecutions: Record[];\n totalUsers: number;\n totalExecutions: number;\n totalCost: number;\n costPerUser: Record[];\n pendingVerifications: number;\n pendingTasks: number;\n}\n\nexport function adminDashboardPage(data: AdminDashboardData): string {\n const {\n currentUser,\n allUsers,\n recentExecutions,\n totalUsers,\n totalExecutions,\n totalCost,\n costPerUser,\n pendingVerifications,\n pendingTasks,\n } = data;\n\n // Stats\n const stats = `\n

${totalUsers}

Total Users

${totalExecutions}

Total Executions

${totalCost.toFixed(2)}

Total Cost

${pendingVerifications}

Pending Verifications

${pendingTasks}

Pending Tasks

\n `;\n\n // Cost per user table\n let costPerUserHtml: string;\n if (costPerUser.length === 0) {\n costPerUserHtml = `

No execution data yet.

`;\n } else {\n const costRows = costPerUser\n .map((row) => {\n const cost = (row.total_cost as number) || 0;\n const execs = (row.execution_count as number) || 0;\n const avgCost = execs > 0 ? cost / execs : 0;\n return `\n ${escapeHtml(row.email as string)}\n ${escapeHtml((row.name as string) || \"—\")}\n ${execs}\n ${cost.toFixed(4)}\n ${avgCost.toFixed(4)}\n `;\n })\n .join(\"\");\n\n costPerUserHtml = `\n

\n \n \n \n \n ${costRows}\n

Email	Name	Executions	Total Cost	Avg Cost

`;\n }\n\n // All users table\n const userRows = allUsers\n .map((u) => {\n const status = u.status as string;\n const admin = (u.is_admin as number) === 1;\n const volume = u.volume_slug\n ? `Active`\n : `None`;\n const statusBadgeHtml =\n status === \"verified\"\n ? `verified`\n : `pending`;\n const adminBadge = admin ? ` admin` : \"\";\n const createdAt = (u.created_at as string) || \"\";\n\n return `\n ${escapeHtml(u.email as string)}\n ${escapeHtml((u.name as string) || \"—\")}\n ${statusBadgeHtml}${adminBadge}\n ${volume}\n ${timeAgo(createdAt)}\n `;\n })\n .join(\"\");\n\n const usersHtml = `\n

\n \n \n \n \n ${userRows}\n

Email	Name	Status	Volume	Joined

`;\n\n // Recent executions (all users)\n let executionsHtml: string;\n if (recentExecutions.length === 0) {\n executionsHtml = `

No executions yet.

`;\n } else {\n const execRows = recentExecutions\n .map((e) => {\n const createdAt = (e.created_at as string) || \"\";\n const elapsed = e.elapsed_ms\n ? `${((e.elapsed_ms as number) / 1000).toFixed(1)}s`\n : \"—\";\n const cost = e.cost_usd\n ? `${(e.cost_usd as number).toFixed(4)}`\n : \"—\";\n const turns = e.turns ?? \"—\";\n const userEmail = (e.user_email as string) || \"—\";\n const errorMsg = e.error_message\n ? `

${escapeHtml((e.error_message as string).slice(0, 60))}

`\n : \"\";\n\n return `\n ${escapeHtml(userEmail)}\n ${triggerBadge(e.trigger_type as string)}\n ${statusBadge(e.status as string)}${errorMsg}\n ${turns}\n ${elapsed}\n ${cost}\n ${timeAgo(createdAt)}\n `;\n })\n .join(\"\");\n\n executionsHtml = `\n

\n \n \n \n \n ${execRows}\n

User	Trigger	Status	Turns	Duration	Cost	When

`;\n }\n\n return baseLayout(\n \"Admin\",\n `\n

🛡️ Admin Dashboard

System-wide overview across all users.

\n\n ${stats}\n\n

Cost Per User

\n ${costPerUserHtml}\n

\n\n

All Users

\n ${usersHtml}\n

\n\n

Recent Executions (all users)

\n ${executionsHtml}\n

\n `,\n currentUser,\n );\n}\n\n// ============ ADMIN USER DETAIL ============\n\ninterface AdminUserDetailData {\n currentUser: UserV2;\n targetUser: Record;\n executions: Record[];\n scheduledTasks: Record[];\n totalExecutions: number;\n totalCost: number;\n pendingEmails: number;\n}\n\nexport function adminUserDetailPage(data: AdminUserDetailData): string {\n const {\n currentUser,\n targetUser,\n executions,\n scheduledTasks,\n totalExecutions,\n totalCost,\n pendingEmails,\n } = data;\n\n const email = targetUser.email as string;\n const name = (targetUser.name as string) || \"—\";\n const status = targetUser.status as string;\n const isAdmin = (targetUser.is_admin as number) === 1;\n const volumeSlug = targetUser.volume_slug as string | null;\n const createdAt = (targetUser.created_at as string) || \"\";\n\n const statusBadgeHtml =\n status === \"verified\"\n ? `verified`\n : `pending`;\n const adminBadge = isAdmin ? ` admin` : \"\";\n const volumeInfo = volumeSlug\n ? `Active ${escapeHtml(volumeSlug)}`\n : `Not provisioned`;\n\n // Stats\n const stats = `\n

${totalExecutions}

Executions

${totalCost.toFixed(2)}

Total Cost

${scheduledTasks.length}

Scheduled Tasks

${pendingEmails}

Pending Emails

\n `;\n\n // User info\n const userInfo = `\n

User Info

\n \n \n \n \n \n \n

Email	${escapeHtml(email)}
Name	${escapeHtml(name)}
Status	${statusBadgeHtml}${adminBadge}
Sandbox Volume	${volumeInfo}
Member since	${createdAt ? new Date(createdAt.endsWith(\"Z\") ? createdAt : createdAt + \"Z\").toLocaleDateString() : \"—\"}

\n `;\n\n // Executions table\n let executionsHtml: string;\n if (executions.length === 0) {\n executionsHtml = `

No executions yet.

`;\n } else {\n const rows = executions\n .map((e) => {\n const cAt = (e.created_at as string) || \"\";\n const elapsed = e.elapsed_ms ? `${((e.elapsed_ms as number) / 1000).toFixed(1)}s` : \"—\";\n const cost = e.cost_usd ? `${(e.cost_usd as number).toFixed(4)}` : \"—\";\n const turns = e.turns ?? \"—\";\n const errorMsg = e.error_message\n ? `

${escapeHtml((e.error_message as string).slice(0, 80))}

`\n : \"\";\n return `\n ${triggerBadge(e.trigger_type as string)}\n ${statusBadge(e.status as string)}${errorMsg}\n ${turns}\n ${elapsed}\n ${cost}\n ${timeAgo(cAt)}\n `;\n })\n .join(\"\");\n\n executionsHtml = `\n

\n \n \n \n \n ${rows}\n

Trigger	Status	Turns	Duration	Cost	When

`;\n }\n\n // Scheduled tasks\n let tasksHtml: string;\n if (scheduledTasks.length === 0) {\n tasksHtml = `

No scheduled tasks.

\n \n \n \n \n ${taskRows}\n

Title	Status	Recurring	Schedule	Next Run

`;\n }\n\n return baseLayout(\n `User: ${email}`,\n `\n

← Back to Admin

${escapeHtml(name !== \"—\" ? name : email)}

${escapeHtml(email)}

\n\n ${stats}\n ${userInfo}\n\n

Executions

\n ${executionsHtml}\n

\n\n

Scheduled Tasks

\n ${tasksHtml}\n

\n `,\n currentUser,\n );\n}\n\n// ============ JSON API ============\n\nexport function jsonResponse(data: unknown, status = 200): Response {\n return new Response(JSON.stringify(data, null, 2), {\n status,\n headers: { \"Content-Type\": \"application/json\" },\n });\n}\n\nexport function htmlResponse(html: string, status = 200, headers?: Record): Response {\n return new Response(html, {\n status,\n headers: { \"Content-Type\": \"text/html; charset=utf-8\", ...headers },\n });\n}\n\nexport function redirect(url: string, headers?: Record): Response {\n return new Response(null, {\n status: 302,\n headers: { Location: url, ...headers },\n });\n}\n",[],[182,275],"e93fbc40-19b4-11f1-b601-42dde27851f2","auth.ts","9919455a-19b0-11f1-aefd-42dde27851f2",120,["D",1772840022310],6,"/**\n * Magic-link authentication helpers.\n *\n * Flow:\n * 1. User enters email on /login\n * 2. We generate a token, store it, email a magic link\n * 3. User clicks link → /auth/verify?token=...\n * 4. We validate the token, create a session, set a cookie\n * 5. Subsequent requests check the session cookie\n */\n\nimport { sqlite } from \"https://esm.town/v/std/sqlite/main.ts\";\nimport { email as sendEmail } from \"https://esm.town/v/std/email\";\nimport { type UserV2, getUserByEmailV2 } from \"../db/schema-v2.ts\";\n\n/** Generate a random hex token */\nfunction generateToken(bytes = 32): string {\n const arr = new Uint8Array(bytes);\n crypto.getRandomValues(arr);\n return Array.from(arr, (b) => b.toString(16).padStart(2, \"0\")).join(\"\");\n}\n\n/** Create a magic link token and email it to the user. */\nexport async function sendMagicLink(\n userEmail: string,\n baseUrl: string,\n): Promise<{ success: boolean; error?: string }> {\n // Search by exact match first, then by LIKE for RFC 5322 formatted addresses\n // (e.g. \"Paul Kinlan \")\n let user = await getUserByEmailV2(userEmail);\n if (!user) {\n // Try matching the email portion inside angle brackets\n const result = await sqlite.execute({\n sql: `SELECT * FROM users WHERE email LIKE ? COLLATE NOCASE`,\n args: [`%<${userEmail.toLowerCase()}>%`],\n });\n if (result.rows.length > 0) {\n const row = result.rows[0] as Record;\n user = {\n id: row.id as number,\n email: row.email as string,\n name: row.name as string | null,\n status: (row.status as \"pending_verification\" | \"verified\") || \"verified\",\n volume_slug: row.volume_slug as string | null,\n volume_id: row.volume_id as string | null,\n volume_region: (row.volume_region as string) || \"ord\",\n sandbox_snapshot: (row.sandbox_snapshot as string) || \"agent-toolchain-v1\",\n sandbox_id: (row.sandbox_id as string) || null,\n created_at: row.created_at as string,\n updated_at: row.updated_at as string,\n };\n }\n }\n\n if (!user) {\n // Don't reveal whether the email exists — just silently succeed\n return { success: true };\n }\n\n const token = generateToken();\n const expiresAt = new Date(Date.now() + 15 * 60 * 1000).toISOString(); // 15 min\n\n await sqlite.execute({\n sql: `INSERT INTO magic_tokens (user_id, token, expires_at) VALUES (?, ?, ?)`,\n args: [user.id, token, expiresAt],\n });\n\n // Extract the bare email address for sending (handles \"Name \" format)\n const bareEmail = extractEmail(user.email);\n const magicUrl = `${baseUrl}/auth/verify?token=${token}`;\n\n try {\n await sendEmail({\n to: bareEmail,\n subject: \"Sign in to Email Agent\",\n replyTo: \"memory-do@valtown.email\",\n text: `Click here to sign in to your Email Agent dashboard:\\n\\n${magicUrl}\\n\\nThis link expires in 15 minutes.\\n\\nIf you didn't request this, just ignore this email.`,\n html: `\n\n\n\n

📧 Email Agent

Click the button below to sign in to your dashboard:

\n Sign In\n

This link expires in 15 minutes.
If you didn't request this, just ignore this email.

\n\n`,\n });\n return { success: true };\n } catch (err) {\n console.error(\"Failed to send magic link:\", err);\n return { success: false, error: \"Failed to send email\" };\n }\n}\n\n/**\n * Extract a bare email address from an RFC 5322 formatted string.\n * e.g. \"Paul Kinlan \" → \"paulkinlan@google.com\"\n * e.g. \"paulkinlan@google.com\" → \"paulkinlan@google.com\"\n */\nexport function extractEmail(raw: string): string {\n const match = raw.match(/<([^>]+)>/);\n if (match) return match[1].trim().toLowerCase();\n return raw.trim().toLowerCase();\n}\n\n/** Verify a magic link token and create a session. Returns a session token or null. */\nexport async function verifyMagicToken(\n token: string,\n): Promise<{ sessionToken: string; user: UserV2 } | null> {\n const now = new Date().toISOString();\n\n const result = await sqlite.execute({\n sql: `SELECT mt.*, u.* FROM magic_tokens mt\n JOIN users u ON mt.user_id = u.id\n WHERE mt.token = ? AND mt.used_at IS NULL AND mt.expires_at > ?`,\n args: [token, now],\n });\n\n if (result.rows.length === 0) return null;\n\n const row = result.rows[0] as Record;\n const userId = row.user_id as number;\n\n // Mark token as used\n await sqlite.execute({\n sql: `UPDATE magic_tokens SET used_at = datetime('now') WHERE token = ?`,\n args: [token],\n });\n\n // Create session (30 days)\n const sessionToken = generateToken();\n const sessionExpires = new Date(\n Date.now() + 30 * 24 * 60 * 60 * 1000,\n ).toISOString();\n\n await sqlite.execute({\n sql: `INSERT INTO sessions (user_id, token, expires_at) VALUES (?, ?, ?)`,\n args: [userId, sessionToken, sessionExpires],\n });\n\n const user: UserV2 = {\n id: userId,\n email: row.email as string,\n name: row.name as string | null,\n status: (row.status as \"pending_verification\" | \"verified\") || \"verified\",\n is_admin: (row.is_admin as number) === 1,\n volume_slug: row.volume_slug as string | null,\n volume_id: row.volume_id as string | null,\n volume_region: (row.volume_region as string) || \"ord\",\n sandbox_snapshot: (row.sandbox_snapshot as string) || \"agent-toolchain-v1\",\n sandbox_id: (row.sandbox_id as string) || null,\n created_at: row.created_at as string,\n updated_at: row.updated_at as string,\n };\n\n return { sessionToken, user };\n}\n\n/** Get the authenticated user from a request's session cookie. Returns null if not authenticated. */\nexport async function getSessionUser(\n req: Request,\n): Promise {\n const cookieHeader = req.headers.get(\"cookie\") || \"\";\n const match = cookieHeader.match(/session=([a-f0-9]+)/);\n if (!match) return null;\n\n const token = match[1];\n const now = new Date().toISOString();\n\n const result = await sqlite.execute({\n sql: `SELECT u.* FROM sessions s\n JOIN users u ON s.user_id = u.id\n WHERE s.token = ? AND s.expires_at > ?`,\n args: [token, now],\n });\n\n if (result.rows.length === 0) return null;\n\n const row = result.rows[0] as Record;\n return {\n id: row.id as number,\n email: row.email as string,\n name: row.name as string | null,\n status: (row.status as \"pending_verification\" | \"verified\") || \"verified\",\n is_admin: (row.is_admin as number) === 1,\n volume_slug: row.volume_slug as string | null,\n volume_id: row.volume_id as string | null,\n volume_region: (row.volume_region as string) || \"ord\",\n sandbox_snapshot: (row.sandbox_snapshot as string) || \"agent-toolchain-v1\",\n sandbox_id: (row.sandbox_id as string) || null,\n created_at: row.created_at as string,\n updated_at: row.updated_at as string,\n };\n}\n\n/** Create a Set-Cookie header value for a session token. */\nexport function sessionCookie(token: string, maxAgeDays = 30): string {\n return `session=${token}; Path=/; HttpOnly; Secure; SameSite=Lax; Max-Age=${maxAgeDays * 86400}`;\n}\n\n/** Create a Set-Cookie header that clears the session. */\nexport function clearSessionCookie(): string {\n return `session=; Path=/; HttpOnly; Secure; SameSite=Lax; Max-Age=0`;\n}\n\n/** Clean up expired tokens and sessions. Call periodically. */\nexport async function cleanupExpiredAuth() {\n await sqlite.batch([\n {\n sql: `DELETE FROM magic_tokens WHERE expires_at < datetime('now')`,\n },\n {\n sql: `DELETE FROM sessions WHERE expires_at < datetime('now')`,\n },\n {\n sql: `DELETE FROM email_verification_tokens WHERE expires_at < datetime('now')`,\n },\n ]);\n}\n",[],[182,283],["D",1772659212112],[214,215,216,217,218,219],[182],"dec6e996-19b4-11f1-b4f7-42dde27851f2","schema-v2.ts","5e83fcba-18bf-11f1-867c-42dde27851f2",118,["D",1772840004730],13,"/**\n * Database schema v2 — Simplified for sandbox-based architecture.\n *\n * With user state living on persistent volumes (memories, todos, ideas as files),\n * SQLite becomes a thin routing/scheduling layer:\n *\n * - **users**: email → user mapping + volume tracking\n * - **scheduled_tasks**: lightweight polling table for the task runner interval\n *\n * The heavy lifting (entries, FTS, entry_links, attachments) is replaced by\n * the filesystem on each user's Deno Sandbox volume, managed by Claude directly.\n */\n\nimport { sqlite } from \"https://esm.town/v/std/sqlite/main.ts\";\n\nexport async function initDatabaseV2() {\n // Core tables\n await sqlite.batch([\n // Users — each unique sender email, with their sandbox volume info\n {\n sql: `CREATE TABLE IF NOT EXISTS users (\n id INTEGER PRIMARY KEY AUTOINCREMENT,\n email TEXT UNIQUE NOT NULL,\n name TEXT,\n status TEXT DEFAULT 'pending_verification',\n volume_slug TEXT UNIQUE,\n volume_id TEXT,\n volume_region TEXT DEFAULT 'ord',\n sandbox_snapshot TEXT DEFAULT 'agent-toolchain-v1',\n sandbox_id TEXT,\n created_at TEXT DEFAULT (datetime('now')),\n updated_at TEXT DEFAULT (datetime('now'))\n )`,\n },\n\n // Pending emails — stored before the user has verified their email address\n {\n sql: `CREATE TABLE IF NOT EXISTS pending_emails (\n id INTEGER PRIMARY KEY AUTOINCREMENT,\n user_id INTEGER NOT NULL,\n subject TEXT,\n body TEXT NOT NULL,\n attachments_json TEXT,\n created_at TEXT DEFAULT (datetime('now')),\n FOREIGN KEY (user_id) REFERENCES users(id)\n )`,\n },\n\n // Email verification tokens — separate from web login magic tokens\n {\n sql: `CREATE TABLE IF NOT EXISTS email_verification_tokens (\n id INTEGER PRIMARY KEY AUTOINCREMENT,\n user_id INTEGER NOT NULL,\n token TEXT UNIQUE NOT NULL,\n expires_at TEXT NOT NULL,\n used_at TEXT,\n created_at TEXT DEFAULT (datetime('now')),\n FOREIGN KEY (user_id) REFERENCES users(id)\n )`,\n },\n\n // Scheduled tasks — just enough for the task runner to poll.\n {\n sql: `CREATE TABLE IF NOT EXISTS scheduled_tasks (\n id INTEGER PRIMARY KEY AUTOINCREMENT,\n user_id INTEGER NOT NULL,\n title TEXT NOT NULL,\n description TEXT NOT NULL,\n next_run_at TEXT NOT NULL,\n schedule_description TEXT,\n is_recurring INTEGER DEFAULT 0,\n status TEXT DEFAULT 'pending',\n created_at TEXT DEFAULT (datetime('now')),\n updated_at TEXT DEFAULT (datetime('now')),\n FOREIGN KEY (user_id) REFERENCES users(id)\n )`,\n },\n\n // Sandbox execution log — track every sandbox boot for cost/debugging\n {\n sql: `CREATE TABLE IF NOT EXISTS sandbox_executions (\n id INTEGER PRIMARY KEY AUTOINCREMENT,\n user_id INTEGER NOT NULL,\n trigger_type TEXT NOT NULL,\n trigger_id TEXT,\n status TEXT DEFAULT 'started',\n turns INTEGER,\n cost_usd REAL,\n elapsed_ms INTEGER,\n error_message TEXT,\n created_at TEXT DEFAULT (datetime('now')),\n completed_at TEXT,\n FOREIGN KEY (user_id) REFERENCES users(id)\n )`,\n },\n\n // Magic link tokens for passwordless login\n {\n sql: `CREATE TABLE IF NOT EXISTS magic_tokens (\n id INTEGER PRIMARY KEY AUTOINCREMENT,\n user_id INTEGER NOT NULL,\n token TEXT UNIQUE NOT NULL,\n expires_at TEXT NOT NULL,\n used_at TEXT,\n created_at TEXT DEFAULT (datetime('now')),\n FOREIGN KEY (user_id) REFERENCES users(id)\n )`,\n },\n\n // Sessions for authenticated users\n {\n sql: `CREATE TABLE IF NOT EXISTS sessions (\n id INTEGER PRIMARY KEY AUTOINCREMENT,\n user_id INTEGER NOT NULL,\n token TEXT UNIQUE NOT NULL,\n expires_at TEXT NOT NULL,\n created_at TEXT DEFAULT (datetime('now')),\n FOREIGN KEY (user_id) REFERENCES users(id)\n )`,\n },\n\n // Indexes\n {\n sql: `CREATE INDEX IF NOT EXISTS idx_users_email ON users(email)`,\n },\n {\n sql: `CREATE INDEX IF NOT EXISTS idx_magic_tokens_token ON magic_tokens(token)`,\n },\n {\n sql: `CREATE INDEX IF NOT EXISTS idx_sessions_token ON sessions(token)`,\n },\n {\n sql: `CREATE INDEX IF NOT EXISTS idx_email_verification_tokens_token ON email_verification_tokens(token)`,\n },\n {\n sql: `CREATE INDEX IF NOT EXISTS idx_pending_emails_user ON pending_emails(user_id)`,\n },\n {\n sql: `CREATE INDEX IF NOT EXISTS idx_scheduled_tasks_due\n ON scheduled_tasks(next_run_at)\n WHERE status = 'pending'`,\n },\n {\n sql: `CREATE INDEX IF NOT EXISTS idx_scheduled_tasks_user\n ON scheduled_tasks(user_id)`,\n },\n {\n sql: `CREATE INDEX IF NOT EXISTS idx_sandbox_executions_user\n ON sandbox_executions(user_id, created_at DESC)`,\n },\n ]);\n\n // Migrate existing users table: add v2 columns if they don't exist.\n // SQLite doesn't have ADD COLUMN IF NOT EXISTS, so we catch errors.\n const newColumns = [\n { name: \"status\", def: \"TEXT DEFAULT 'verified'\" }, // existing users default to verified\n { name: \"is_admin\", def: \"INTEGER DEFAULT 0\" },\n { name: \"volume_slug\", def: \"TEXT\" },\n { name: \"volume_id\", def: \"TEXT\" },\n { name: \"volume_region\", def: \"TEXT DEFAULT 'ord'\" },\n { name: \"sandbox_snapshot\", def: \"TEXT DEFAULT 'agent-toolchain-v1'\" },\n { name: \"sandbox_id\", def: \"TEXT\" },\n { name: \"updated_at\", def: \"TEXT DEFAULT (datetime('now'))\" },\n ];\n for (const col of newColumns) {\n try {\n await sqlite.execute({\n sql: `ALTER TABLE users ADD COLUMN ${col.name} ${col.def}`,\n args: [],\n });\n console.log(` ✅ Added column users.${col.name}`);\n } catch {\n // Column already exists — that's fine\n }\n }\n\n console.log(\"💾 Database schema v2 initialized\");\n}\n\n// ============ HELPERS ============\n\n/**\n * Extract a bare email address from an RFC 5322 formatted string.\n * e.g. \"Paul Kinlan \" → \"paulkinlan@google.com\"\n * e.g. \"paulkinlan@google.com\" → \"paulkinlan@google.com\"\n */\nexport function extractEmail(raw: string): string {\n const match = raw.match(/<([^>]+)>/);\n if (match) return match[1].trim().toLowerCase();\n return raw.trim().toLowerCase();\n}\n\n/**\n * Extract the display name from an RFC 5322 formatted string.\n * e.g. \"Paul Kinlan \" → \"Paul Kinlan\"\n * e.g. \"paulkinlan@google.com\" → null\n */\nfunction extractName(raw: string): string | null {\n const match = raw.match(/^(.+?)\\s*<[^>]+>$/);\n if (match) return match[1].trim();\n return null;\n}\n\n// ============ USER QUERIES ============\n\nexport interface UserV2 {\n id: number;\n email: string;\n name: string | null;\n status: \"pending_verification\" | \"verified\";\n is_admin: boolean;\n volume_slug: string | null;\n volume_id: string | null;\n volume_region: string;\n sandbox_snapshot: string;\n sandbox_id: string | null;\n created_at: string;\n updated_at: string;\n}\n\n/**\n * Get or create a user. New users start as 'pending_verification'.\n * Returns { user, isNew } so callers can decide whether to send verification.\n */\nexport async function getOrCreateUserV2(\n rawEmail: string,\n name?: string,\n): Promise<{ user: UserV2; isNew: boolean }> {\n // Normalize: extract bare email from \"Name \" format\n const bareEmail = extractEmail(rawEmail);\n const derivedName = name || extractName(rawEmail);\n\n const existing = await sqlite.execute({\n sql: `SELECT * FROM users WHERE email = ?`,\n args: [bareEmail],\n });\n\n if (existing.rows.length > 0) {\n const row = existing.rows[0] as Record;\n if (derivedName && row.name !== derivedName) {\n await sqlite.execute({\n sql: `UPDATE users SET name = ?, updated_at = datetime('now') WHERE id = ?`,\n args: [derivedName, row.id as number],\n });\n }\n return { user: parseUserRow(row), isNew: false };\n }\n\n const result = await sqlite.execute({\n sql: `INSERT INTO users (email, name, status) VALUES (?, ?, 'pending_verification')`,\n args: [bareEmail, derivedName || null],\n });\n\n return {\n user: {\n id: Number(result.lastInsertRowid),\n email: bareEmail,\n name: derivedName || null,\n status: \"pending_verification\",\n is_admin: false,\n volume_slug: null,\n volume_id: null,\n volume_region: \"ord\",\n sandbox_snapshot: \"agent-toolchain-v1\",\n sandbox_id: null,\n created_at: new Date().toISOString(),\n updated_at: new Date().toISOString(),\n },\n isNew: true,\n };\n}\n\n/** Mark a user as verified. */\nexport async function verifyUser(userId: number) {\n await sqlite.execute({\n sql: `UPDATE users SET status = 'verified', updated_at = datetime('now') WHERE id = ?`,\n args: [userId],\n });\n}\n\n// ============ PENDING EMAIL QUERIES ============\n\nexport interface PendingEmail {\n id: number;\n user_id: number;\n subject: string | null;\n body: string;\n attachments_json: string | null;\n created_at: string;\n}\n\n/** Store an email for later processing (pre-verification). */\nexport async function storePendingEmail(opts: {\n userId: number;\n subject: string | null;\n body: string;\n attachmentsJson?: string;\n}): Promise {\n const result = await sqlite.execute({\n sql: `INSERT INTO pending_emails (user_id, subject, body, attachments_json) VALUES (?, ?, ?, ?)`,\n args: [opts.userId, opts.subject, opts.body, opts.attachmentsJson || null],\n });\n return Number(result.lastInsertRowid);\n}\n\n/** Get all pending emails for a user. */\nexport async function getPendingEmails(userId: number): Promise {\n const result = await sqlite.execute({\n sql: `SELECT * FROM pending_emails WHERE user_id = ? ORDER BY created_at ASC`,\n args: [userId],\n });\n return (result.rows as Record[]).map((row) => ({\n id: row.id as number,\n user_id: row.user_id as number,\n subject: row.subject as string | null,\n body: row.body as string,\n attachments_json: row.attachments_json as string | null,\n created_at: row.created_at as string,\n }));\n}\n\n/** Delete all pending emails for a user (after they've been processed). */\nexport async function deletePendingEmails(userId: number) {\n await sqlite.execute({\n sql: `DELETE FROM pending_emails WHERE user_id = ?`,\n args: [userId],\n });\n}\n\n// ============ EMAIL VERIFICATION TOKEN QUERIES ============\n\n/** Create an email verification token. Returns the token string. */\nexport async function createEmailVerificationToken(userId: number): Promise {\n const arr = new Uint8Array(32);\n crypto.getRandomValues(arr);\n const token = Array.from(arr, (b) => b.toString(16).padStart(2, \"0\")).join(\"\");\n const expiresAt = new Date(Date.now() + 24 * 60 * 60 * 1000).toISOString(); // 24 hours\n\n await sqlite.execute({\n sql: `INSERT INTO email_verification_tokens (user_id, token, expires_at) VALUES (?, ?, ?)`,\n args: [userId, token, expiresAt],\n });\n\n return token;\n}\n\n/** Verify an email verification token. Returns the user ID if valid, null otherwise. */\nexport async function verifyEmailToken(token: string): Promise {\n const now = new Date().toISOString();\n\n const result = await sqlite.execute({\n sql: `SELECT * FROM email_verification_tokens\n WHERE token = ? AND used_at IS NULL AND expires_at > ?`,\n args: [token, now],\n });\n\n if (result.rows.length === 0) return null;\n\n const row = result.rows[0] as Record;\n const userId = row.user_id as number;\n\n // Mark token as used\n await sqlite.execute({\n sql: `UPDATE email_verification_tokens SET used_at = datetime('now') WHERE token = ?`,\n args: [token],\n });\n\n return userId;\n}\n\nexport async function getUserV2(userId: number): Promise {\n const result = await sqlite.execute({\n sql: `SELECT * FROM users WHERE id = ?`,\n args: [userId],\n });\n if (result.rows.length === 0) return undefined;\n return parseUserRow(result.rows[0] as Record);\n}\n\nexport async function getUserByEmailV2(email: string): Promise {\n const result = await sqlite.execute({\n sql: `SELECT * FROM users WHERE email = ?`,\n args: [email.toLowerCase()],\n });\n if (result.rows.length === 0) return undefined;\n return parseUserRow(result.rows[0] as Record);\n}\n\nexport async function updateUserVolume(\n userId: number,\n volumeSlug: string,\n volumeId: string,\n region: string,\n) {\n await sqlite.execute({\n sql: `UPDATE users SET volume_slug = ?, volume_id = ?, volume_region = ?, updated_at = datetime('now') WHERE id = ?`,\n args: [volumeSlug, volumeId, region, userId],\n });\n}\n\nexport async function updateUserSandboxId(\n userId: number,\n sandboxId: string | null,\n) {\n await sqlite.execute({\n sql: `UPDATE users SET sandbox_id = ?, updated_at = datetime('now') WHERE id = ?`,\n args: [sandboxId, userId],\n });\n}\n\n// ============ SCHEDULED TASK QUERIES ============\n\nexport interface ScheduledTask {\n id: number;\n user_id: number;\n title: string;\n description: string;\n next_run_at: string;\n schedule_description: string | null;\n is_recurring: boolean;\n status: string;\n created_at: string;\n updated_at: string;\n}\n\nexport async function addScheduledTask(opts: {\n userId: number;\n title: string;\n description: string;\n nextRunAt: string;\n scheduleDescription?: string;\n isRecurring?: boolean;\n}): Promise {\n const result = await sqlite.execute({\n sql: `INSERT INTO scheduled_tasks (user_id, title, description, next_run_at, schedule_description, is_recurring)\n VALUES (?, ?, ?, ?, ?, ?)`,\n args: [\n opts.userId,\n opts.title,\n opts.description,\n opts.nextRunAt,\n opts.scheduleDescription || null,\n opts.isRecurring ? 1 : 0,\n ],\n });\n return Number(result.lastInsertRowid);\n}\n\nexport async function getScheduledTasksDue(): Promise<\n (ScheduledTask & { user_email: string; user_name: string | null; volume_slug: string | null })[\n ]\n> {\n const now = new Date().toISOString().replace(\"T\", \" \").replace(\"Z\", \"\");\n const result = await sqlite.execute({\n sql: `SELECT st.*, u.email as user_email, u.name as user_name, u.volume_slug\n FROM scheduled_tasks st\n JOIN users u ON st.user_id = u.id\n WHERE st.status = 'pending'\n AND REPLACE(REPLACE(st.next_run_at, 'T', ' '), 'Z', '') <= ?\n ORDER BY st.next_run_at ASC\n LIMIT 20`,\n args: [now],\n });\n return (result.rows as Record[]).map((row) => ({\n ...parseScheduledTaskRow(row),\n user_email: row.user_email as string,\n user_name: row.user_name as string | null,\n volume_slug: row.volume_slug as string | null,\n }));\n}\n\nexport async function updateScheduledTaskStatus(\n taskId: number,\n status: string,\n) {\n await sqlite.execute({\n sql: `UPDATE scheduled_tasks SET status = ?, updated_at = datetime('now') WHERE id = ?`,\n args: [status, taskId],\n });\n}\n\nexport async function rescheduleTask(taskId: number, nextRunAt: string) {\n await sqlite.execute({\n sql: `UPDATE scheduled_tasks SET status = 'pending', next_run_at = ?, updated_at = datetime('now') WHERE id = ?`,\n args: [nextRunAt, taskId],\n });\n}\n\n// ============ SANDBOX EXECUTION LOG ============\n\nexport async function logSandboxExecution(opts: {\n userId: number;\n triggerType: \"email\" | \"scheduled_task\" | \"admin\";\n triggerId?: string;\n}): Promise {\n const result = await sqlite.execute({\n sql: `INSERT INTO sandbox_executions (user_id, trigger_type, trigger_id)\n VALUES (?, ?, ?)`,\n args: [opts.userId, opts.triggerType, opts.triggerId || null],\n });\n return Number(result.lastInsertRowid);\n}\n\nexport async function completeSandboxExecution(\n executionId: number,\n result: {\n status: \"success\" | \"failure\" | \"timeout\";\n turns?: number;\n costUsd?: number;\n elapsedMs?: number;\n errorMessage?: string;\n },\n) {\n await sqlite.execute({\n sql: `UPDATE sandbox_executions\n SET status = ?, turns = ?, cost_usd = ?, elapsed_ms = ?, error_message = ?, completed_at = datetime('now')\n WHERE id = ?`,\n args: [\n result.status,\n result.turns || null,\n result.costUsd || null,\n result.elapsedMs || null,\n result.errorMessage || null,\n executionId,\n ],\n });\n}\n\n// ============ HELPERS ============\n\nfunction parseUserRow(row: Record): UserV2 {\n return {\n id: row.id as number,\n email: row.email as string,\n name: row.name as string | null,\n status: (row.status as \"pending_verification\" | \"verified\") || \"verified\",\n is_admin: (row.is_admin as number) === 1,\n volume_slug: row.volume_slug as string | null,\n volume_id: row.volume_id as string | null,\n volume_region: (row.volume_region as string) || \"ord\",\n sandbox_snapshot: (row.sandbox_snapshot as string) || \"agent-toolchain-v1\",\n sandbox_id: (row.sandbox_id as string) || null,\n created_at: row.created_at as string,\n updated_at: row.updated_at as string,\n };\n}\n\nfunction parseScheduledTaskRow(row: Record): ScheduledTask {\n return {\n id: row.id as number,\n user_id: row.user_id as number,\n title: row.title as string,\n description: row.description as string,\n next_run_at: row.next_run_at as string,\n schedule_description: row.schedule_description as string | null,\n is_recurring: (row.is_recurring as number) === 1,\n status: row.status as string,\n created_at: row.created_at as string,\n updated_at: row.updated_at as string,\n };\n}\n",[],[190,295],["D",1772659211984],[212],[190],"e4594830-18c5-11f1-9e0c-42dde27851f2","test-orchestrator.ts","5e838f1e-18bf-11f1-867c-42dde27851f2",61,["D",1772737364373],"/**\n * Test: Sandbox with @anthropic-ai/claude-agent-sdk\n * Tests the corrected package and API.\n */\nimport { Sandbox } from \"jsr:@deno/sandbox\";\nimport { initDatabaseV2 } from \"../db/schema-v2.ts\";\nimport {\n AGENT_LOOP_PACKAGE_JSON,\n AGENT_LOOP_SOURCE,\n} from \"../lib/sandbox-agent-loop.ts\";\n\nconst USER_HOME = \"/home/app\";\nconst AGENT_DIR = \"/home/app/agent\";\n\nconsole.log(\"🧪 Testing with @anthropic-ai/claude-agent-sdk\\n\");\n\nawait initDatabaseV2();\n\nconsole.log(\"📦 Creating sandbox...\");\nawait using sandbox = await Sandbox.create({\n env: {\n HOME: USER_HOME,\n ANTHROPIC_API_KEY: Deno.env.get(\"ANTHROPIC_API_KEY\") || \"\",\n AGENT_MODEL: \"claude-sonnet-4-6\",\n AGENT_MAX_TURNS: \"5\",\n AGENT_CYCLE_TIMEOUT_MS: \"60000\",\n },\n});\nconsole.log(\" ✅ Sandbox created\\n\");\n\n// Install\nconsole.log(\"📦 Installing agent-loop...\");\nawait sandbox.sh`mkdir -p ${AGENT_DIR}`;\nawait sandbox.fs.writeTextFile(\n `${AGENT_DIR}/agent-loop.mjs`,\n AGENT_LOOP_SOURCE,\n);\nawait sandbox.fs.writeTextFile(\n `${AGENT_DIR}/package.json`,\n AGENT_LOOP_PACKAGE_JSON,\n);\nawait sandbox.sh`cd ${AGENT_DIR} && npm install 2>&1`;\nconsole.log(\" ✅ Installed\\n\");\n\n// Check what got installed\nconsole.log(\"=== 1. What's in node_modules/@anthropic-ai/ ===\");\nconst aiPkgs = await sandbox\n .sh`ls ${AGENT_DIR}/node_modules/@anthropic-ai/ 2>/dev/null`.text();\nconsole.log(aiPkgs);\n\nconsole.log(\"\\n=== 2. claude-agent-sdk package.json ===\");\nconst rawPkg = await sandbox\n .sh`cat ${AGENT_DIR}/node_modules/@anthropic-ai/claude-agent-sdk/package.json`\n .text();\nconst pkgData = JSON.parse(rawPkg);\nconsole.log(\" name:\", pkgData.name);\nconsole.log(\" type:\", pkgData.type);\nconsole.log(\" main:\", pkgData.main);\nconsole.log(\" exports:\", JSON.stringify(pkgData.exports)?.substring(0, 500));\nconsole.log(\" bin:\", JSON.stringify(pkgData.bin)?.substring(0, 200));\n\nconsole.log(\"\\n=== 3. Package files ===\");\nconst pkgFiles = await sandbox\n .sh`ls ${AGENT_DIR}/node_modules/@anthropic-ai/claude-agent-sdk/`.text();\nconsole.log(pkgFiles);\n\n// Test the import\nconsole.log(\"\\n=== 4. Test ESM import ===\");\nawait sandbox.fs.writeTextFile(\n `${AGENT_DIR}/test-import.mjs`,\n 'try {\\n const pkg = await import(\"@anthropic-ai/claude-agent-sdk\");\\n console.log(\"keys:\", Object.keys(pkg).join(\", \"));\\n console.log(\"query:\", typeof pkg.query);\\n} catch(e) {\\n console.log(\"error:\", e.message);\\n console.log(\"stack:\", e.stack?.substring(0, 800));\\n}\\n',\n);\nawait sandbox.fs.writeTextFile(\n \"/tmp/test-import.sh\",\n `cd ${AGENT_DIR} && node test-import.mjs > /tmp/import.log 2>&1; echo \"EXIT=\\$?\" >> /tmp/import.log`,\n);\nawait sandbox.sh`bash /tmp/test-import.sh`;\nconsole.log(await sandbox.fs.readTextFile(\"/tmp/import.log\"));\n\n// Run the full agent-loop\nconsole.log(\"\\n=== 5. Run agent-loop ===\");\nawait sandbox\n .sh`mkdir -p ${USER_HOME}/.mail/inbox ${USER_HOME}/.mail/outbox ${USER_HOME}/.agent-results ${USER_HOME}/memories`;\nawait sandbox.fs.writeTextFile(\n `${USER_HOME}/CLAUDE.md`,\n \"You are a test agent. Say hello and list what tools you have.\",\n);\nawait sandbox.fs.writeTextFile(`${USER_HOME}/TODO.md`, \"\");\n\nconst taskId = Date.now();\nawait sandbox.fs.writeTextFile(\n `${USER_HOME}/.mail/inbox/task-${taskId}.md`,\n \"# Test\\nSay hello briefly.\\n\\n## Instructions\\nWrite response to: `/home/app/.mail/outbox/response-\" +\n taskId + \".md`\\n\",\n);\n\nawait sandbox.fs.writeTextFile(\n \"/tmp/run-agent.sh\",\n `cd ${USER_HOME} && node ${AGENT_DIR}/agent-loop.mjs > /tmp/agent.log 2>&1; echo \"EXIT=\\$?\" >> /tmp/agent.log`,\n);\n\nconsole.log(\" Running... (may take 30s-2min)\");\nawait sandbox.sh`bash /tmp/run-agent.sh`;\nconst agentLog = await sandbox.fs.readTextFile(\"/tmp/agent.log\");\nconsole.log(agentLog);\n\n// Check results\nconsole.log(\"\\n=== 6. Post state ===\");\nconsole.log(\n \"Heartbeat:\",\n await sandbox\n .sh`cat ${USER_HOME}/.agent-heartbeat 2>/dev/null || echo \"(none)\"`.text(),\n);\nconsole.log(\n \"Outbox:\",\n await sandbox.sh`ls ${USER_HOME}/.mail/outbox/ 2>/dev/null || echo \"(empty)\"`\n .text(),\n);\nconsole.log(\n \"Results:\",\n await sandbox\n .sh`ls ${USER_HOME}/.agent-results/ 2>/dev/null || echo \"(empty)\"`.text(),\n);\n\n// Read response if it exists\ntry {\n const response = await sandbox.fs.readTextFile(\n `${USER_HOME}/.mail/outbox/response-${taskId}.md`,\n );\n console.log(\"\\n📝 Response:\\n\" + response);\n} catch {\n console.log(\"\\n(No response file)\");\n}\n\n// Read cycle result\ntry {\n const latestResult = await sandbox\n .sh`ls -t ${USER_HOME}/.agent-results/ | head -1`.text();\n if (latestResult.trim()) {\n const result = await sandbox.fs.readTextFile(\n `${USER_HOME}/.agent-results/${latestResult.trim()}`,\n );\n console.log(\"\\n📊 Cycle result:\\n\" + result);\n }\n} catch {\n console.log(\"(No cycle result)\");\n}\n\nconsole.log(\"\\n✅ Done\");",[],[195,307],"5e834680-18bf-11f1-867c-42dde27851f2","build-snapshot.ts","5e83473e-18bf-11f1-867c-42dde27851f2",["D",1772734562850],"/**\n * Admin: Build Toolchain Snapshot\n *\n * Sets up a sandbox with the agent-loop and verifies all tools.\n *\n * Base image already provides: git, ssh, curl, node, npm, npx, deno, python3, sudo\n * We just need to: install a few extras, set up agent-loop, configure ssh/git.\n *\n * Key facts about the Deno Sandbox base image:\n * - OS: Debian 13 (trixie), x86_64\n * - User: app (home: /home/app)\n * - sudo works for apt-get\n * - Pre-installed: git, ssh, curl, node, npm, npx, deno, python3\n */\n\nimport { Sandbox } from \"jsr:@deno/sandbox\";\nimport { AGENT_LOOP_SOURCE, AGENT_LOOP_PACKAGE_JSON } from \"../lib/sandbox-agent-loop.ts\";\n\nconst AGENT_DIR = \"/home/app/agent\";\n\nexport async function buildToolchainSnapshot() {\n console.log(\"🔧 Building toolchain snapshot...\");\n\n console.log(\"📦 Creating sandbox...\");\n await using sandbox = await Sandbox.create();\n\n // 1. Install a few extra system packages\n console.log(\"\\n📦 Installing extra packages...\");\n await sandbox.sh`sudo apt-get update -y`;\n await sandbox.sh`sudo apt-get install -y jq tree wget unzip`;\n console.log(\" ✅ Extra packages installed\");\n\n // 2. Verify pre-installed tools\n console.log(\"\\n🔍 Verifying tools:\");\n const checks = [\n { name: \"git\", cmd: \"git --version\" },\n { name: \"node\", cmd: \"node --version\" },\n { name: \"npm\", cmd: \"npm --version\" },\n { name: \"deno\", cmd: \"deno --version\" },\n { name: \"python3\", cmd: \"python3 --version\" },\n { name: \"ssh\", cmd: \"ssh -V\" },\n { name: \"curl\", cmd: \"curl --version\" },\n { name: \"jq\", cmd: \"jq --version\" },\n { name: \"tree\", cmd: \"tree --version\" },\n ];\n for (const { name, cmd } of checks) {\n try {\n const output = await sandbox.sh`bash -c ${cmd}`.text();\n console.log(` ✅ ${name}: ${output.trim().split(\"\\n\")[0]}`);\n } catch {\n console.log(` ❌ ${name}: not available`);\n }\n }\n\n // 3. Set up agent-loop in user's home (no sudo needed)\n console.log(`\\n📦 Setting up agent-loop in ${AGENT_DIR}...`);\n await sandbox.sh`mkdir -p ${AGENT_DIR}`;\n await sandbox.fs.writeTextFile(`${AGENT_DIR}/agent-loop.mjs`, AGENT_LOOP_SOURCE);\n await sandbox.fs.writeTextFile(`${AGENT_DIR}/package.json`, AGENT_LOOP_PACKAGE_JSON);\n\n console.log(\" Installing npm dependencies...\");\n const installOutput = await sandbox.sh`cd ${AGENT_DIR} && npm install 2>&1`.text();\n // Show last few lines of install output\n const lines = installOutput.trim().split(\"\\n\");\n for (const line of lines.slice(-5)) {\n console.log(` ${line}`);\n }\n console.log(\" ✅ npm install complete\");\n\n // 4. Set up SSH config\n console.log(\"\\n🔧 Setting up SSH config...\");\n await sandbox.sh`mkdir -p /home/app/.ssh`;\n await sandbox.sh`chmod 700 /home/app/.ssh`;\n await sandbox.fs.writeTextFile(\n \"/home/app/.ssh/config\",\n \"Host github.com\\n StrictHostKeyChecking accept-new\\n IdentityFile ~/.ssh/id_ed25519\\n\",\n );\n await sandbox.sh`chmod 600 /home/app/.ssh/config`;\n console.log(\" ✅ SSH config written\");\n\n // 5. Set up git config\n console.log(\"🔧 Setting up git config...\");\n await sandbox.fs.writeTextFile(\n \"/home/app/.gitconfig\",\n \"[init]\\n defaultBranch = main\\n[safe]\\n directory = *\\n\",\n );\n console.log(\" ✅ Git config written\");\n\n // 6. Final verification\n console.log(\"\\n🧪 Final verification:\");\n\n const agentFiles = await sandbox.sh`ls -la ${AGENT_DIR}/`.text();\n console.log(agentFiles);\n\n const nodeModules = await sandbox.sh`bash -c \"ls ${AGENT_DIR}/node_modules/ 2>/dev/null | head -20\"`.text();\n console.log(`node_modules: ${nodeModules}`);\n\n // Quick test: can node parse the agent-loop?\n try {\n const parseCheck = await sandbox.sh`node --check ${AGENT_DIR}/agent-loop.mjs 2>&1 || echo \"parse failed\"`.text();\n console.log(` ✅ agent-loop.mjs syntax check: ${parseCheck.trim() || \"OK\"}`);\n } catch (e) {\n console.log(` ⚠️ agent-loop.mjs parse check: ${e}`);\n }\n\n // Show home directory structure\n const homeTree = await sandbox.sh`tree -L 2 /home/app/ -a`.text();\n console.log(`\\n📂 Home directory:\\n${homeTree}`);\n\n console.log(`\\n✅ Toolchain setup complete!`);\n console.log(`\\nKey info for orchestrator:`);\n console.log(` User: app`);\n console.log(` Home: /home/app`);\n console.log(` Agent dir: ${AGENT_DIR}`);\n console.log(` Agent loop: ${AGENT_DIR}/agent-loop.mjs`);\n}\n\n// Run when Val Town executes this script\nawait buildToolchainSnapshot();\n",[],[195,315],["D",1772734562850],[209,210],[195],["D",1772647273772],["D",1789516910456],{"_11":12,"_97":93},{"_11":12,"_90":91,"_92":93,"_52":53,"_94":95,"_96":46},{"_30":42,"_13":41,"_21":33,"_28":29,"_32":-5,"_34":-5,"_85":-5,"_86":-5,"_87":-5,"_88":14,"_89":12},"ctx",{"_78":-5,"_120":362,"_363":364,"_365":366,"_367":46,"_368":46,"_369":46,"_370":118,"_371":-5,"_372":-7,"_373":-7,"_374":186,"_375":46,"_376":46,"_377":-5,"_378":46,"_379":46,"_380":46,"_381":-5},"ENV",{"_333":334,"_335":336,"_337":41,"_338":339,"_340":341,"_342":343,"_344":345,"_346":347,"_348":349,"_350":351,"_352":353,"_354":355,"_356":357,"_358":359,"_360":361},"REACT_APP_SENTRY_DSN","https://586020d86f05489db63a554690acf1e9@o1398359.ingest.sentry.io/6724499","RENDER_GIT_COMMIT","e87eb04e707cf64e7e532c8e8ec395f209e17f15","RENDER_GIT_BRANCH","POSTHOG_API_TOKEN","phc_nmoyuAkStEZuUp69rYlg1DZgrIt19Gx0Y4Hwf8tGgJP","POSTHOG_API_HOST","https://wart.val.town","WEB_API_TEMPLATE","https://{val}.web.val.run","APP_URL","https://www.val.town","MODULE_URL","https://esm.town","VANITY_WEB_API_TEMPLATE","https://{label}.val.run/","DENOLS_URL","https://lsp.val.town","NODE_ENV","production","CLOUDFLARE_TURNSTILE_SITE_KEY","0x4AAAAAACtaIZJH7s9rBMZ6","GOOGLE_CLIENT_ID","875989696899-jtsqkhlbdv5jrlsk8631p7ipjg3c6o62.apps.googleusercontent.com","STRIPE_PUBLISHABLE_KEY","pk_live_51N2zvWEpZTTLipIkVEwXrUG9AT0yeOOEYxszarZT1jpsZV58Ly6wlg3SltZpHN26CUalOBmUSY7adQRQiSkCZz2500qSXNkf8Y","API_URL","https://api.val.town",[],"preferencesKeybindings","default","preferencesTheme","light","preferencesFixers","preferencesCopilot","preferencesSelectionAi","preferencesCloseBrackets","preferencesChatPanelState","preferencesDprint","preferencesDenofmt","notificationCount","flagDenols","isBusinessOrgMember","businessOrgHandle","isBusinessOrgAdmin","isHigherTownieTier","navCollapsed","currentOrgId"]

Branches2

Branches
2