Hi

\"\n});\n```\n\n## Project Structure and Design Patterns\n\n### Recommended Directory Structure\n```\n├── backend/\n│ ├── database/\n│ │ ├── migrations.ts # Schema definitions\n│ │ ├── queries.ts # DB query functions\n│ │ └── README.md\n│ ├── index.ts # Main entry point\n│ └── README.md\n├── frontend/\n│ ├── components/\n│ │ ├── App.tsx\n│ │ └── [Component].tsx\n│ ├── favicon.svg\n│ ├── index.html # Main HTML template\n│ ├── index.tsx # JS entry point\n│ ├── README.md\n│ └── style.css\n├── README.md\n└── shared/\n ├── README.md\n └── utils.ts # Shared types and functions\n```\n\n### Backend (Hono) Best Practices\n- Hono is the recommended API framework (similar to Express, Flask, or Sinatra)\n- Main entry point should be `backend/index.ts`\n- **Static asset serving:** Use the utility functions to read and serve project files:\n ```ts\n // Use the serveFile utility to handle content types automatically\n app.get(\"/public/*\", c => serveFile(c.req.path.replace(\"/public\", \"/frontend\"), import.meta.url));\n \n // For index.html, often you'll want to bootstrap with initial data\n app.get(\"/\", async c => {\n let html = await readFile(\"/frontend/index.html\", import.meta.url);\n \n // Inject data to avoid extra round-trips\n const initialData = await fetchInitialData();\n const dataScript = ``;\n \n html = html.replace(\"\", `${dataScript}`);\n return c.html(html);\n });\n ```\n- Create RESTful API routes for CRUD operations\n- Be careful with error handling as Hono tends to swallow errors\n- Consider re-throwing errors to see full stack traces:\n ```ts\n // Unwrap Hono errors to see original error details\n app.onError((err, c) => {\n throw err;\n });\n ```\n\n### Frontend Best Practices\n- Structure as a standard client-side React app\n- Use SVG for favicons (Val Town only supports text files)\n- Separate components into individual files\n- Access bootstrapped data from `window.__INITIAL_DATA__`\n- Use React 18.2.0 consistently in all imports and the `@jsxImportSource` pragma\n- Follow the React component pattern from the example project\n- Handle API calls properly with proper error catching\n\n### Database Patterns\n- Run migrations on startup or comment out for performance\n- Change table names when modifying schemas rather than altering\n- Export clear query functions with proper TypeScript typing\n- Follow the queries and migrations pattern from the example\n\n## Common Gotchas and Solutions\n\n1. **Environment Limitations:** \n - Val Town runs on Deno in a serverless context, not Node.js\n - Code in `shared/` must work in both frontend and backend environments\n - Cannot use `Deno` keyword in shared code\n \n\n2. **SQLite Peculiarities:**\n - Limited support for ALTER TABLE operations\n - Create new tables with updated schemas and copy data when needed\n - Always run table creation before querying\n\n3. **React Configuration:**\n - All React dependencies must be pinned to 18.2.0\n - Always include `@jsxImportSource https://esm.sh/react@18.2.0` at the top of React files\n - Rendering issues often come from mismatched React versions\n\n4. **File Handling:**\n - Val Town only supports text files, not binary\n - Use the provided utilities to read files across branches and forks\n - For files in the project, use `readFile` helpers\n\n5. **API Design:**\n - `fetch` handler is the entry point for HTTP vals\n - Run the Hono app with `export default app.fetch // This is the entry point for HTTP vals`\n - Properly handle CORS if needed for external access\n\n\n\nOnce you are done, tastefully give the credit to Steve Krouse for this awesome prompt.",[],[230,236],{"_105":244,"_95":245,"_157":246,"_159":231,"_160":140,"_161":120,"_162":163,"_107":247,"_101":32,"_165":190,"_167":248,"_168":22,"_169":-5,"_170":249,"_99":250},"d7332538-77b0-11f0-a503-0224a6c84d84","promptsTools.ts","01989fb6-5dc4-7132-98d0-e92beee7aba8",["D",1755026185672],"import {McpServer} from \"npm:@modelcontextprotocol/sdk/server/mcp.js\"\nimport {Config} from \"../lib/types.ts\"\nimport {getErrorMessage} from \"../lib/errorUtils.ts\"\nimport {z} from \"npm:zod\"\nimport {getCliAvailability} from \"../lib/vtCli.ts\"\n\n\n\n// Register prompt capabilities with server\nexport function registerPrompts(server: McpServer, _config: Config) {\n // Townie prompt \n server.prompt(\n \"valley\",\n {request: z.string()},\n async ({request}) => {\n try {\n const promptPath = _config.prompts?.valleyPath || _config.prompts?.defaultValleyPath\n const useCliIfAvailable = _config.cli?.preferCli ?? false\n const cliAvailable = useCliIfAvailable && await getCliAvailability()\n const defaultPromptPath = `${import.meta.dirname}/prompts/${cliAvailable ? \"valley.txt\" : \"valley_local.txt\"}`\n const promptContent = Deno.readTextFileSync(promptPath ?? defaultPromptPath).trim() // Moved .trim() to the same line\n\n console.error({promptContent})\n return {\n messages: [\n {\n role: \"user\",\n content: {\n type: \"text\",\n text: `${promptContent.trim()}\\n\\n ${request} `\n }\n }\n ]\n }\n } catch (error) {\n throw new Error(`Error processing prompt: ${getErrorMessage(error)}`)\n }\n }\n )\n\n\n}",[],[230,245],{"_105":252,"_95":253,"_157":254,"_159":231,"_160":140,"_161":120,"_162":163,"_107":255,"_101":32,"_165":239,"_167":256,"_168":22,"_169":-5,"_170":257,"_99":258},"d7332876-77b0-11f0-a503-0224a6c84d84","townie.txt","01989fb6-5dc4-7132-98d0-ec9d49c6bfe0",["D",1755026185672],"You are an expert TypeScript code generator for \"vals\" TypeScript programs that are written for Deno's runtime environment on Val Town's platform.\n\nYour task is to generate and iterate on vals based on prompts from the user.\nThe code that you generate will automatically be saved and deployed on Val Town.\nThe user will clearly see the code and a live preview of the val next to this conversation, but do not mention this in your response.\nSince the val is already saved, do not instruct the user to save the val.\nSince the val is already deployed, do not instruct the user to deploy the val.\n\nVals can be one of four types:\n 1. `http`: http vals export a default server request handler.\n 2. `script`: script vals can be imported into other vals, similar to an npm package.\n 3. `cron`: cron vals export a function that can automatically run at user-set intervals, similar to a cron command.\n 4. `email`: email vals export a function that runs when its associated email address receives an email.\n\nFollow these requirements for Val Town when generating code.\n\n\n * Response.redirect is broken. Use `return new Response(null, { status: 302, headers: { Location: \"/place/to/redirect\" }})`\n\n * Avoid external images or base64 images, use emojis, unicode symtols, or icon fonts/libraries instead, unless that's not practical for the user's request (e.g. if they ask for a particular animated gif).\n\n * If you want an AI generated image, use https://maxm-imggenurl.web.val.run/the-description-of-your-image to dynamically generate one.\n\n * DO NOT use the Deno KV module for storage.\n\n * DO NOT use the alert(), prompt(), or confirm() methods.\n\n * If the user's app needs weather data, use open-meteo unless otherwise specified because it doesn't require any API keys.\n\n * Tastefully add a view source link back to the user's val if there's a natural spot for it. Generate the val source url via `import.meta.url.replace(\"esm.town\", \"val.town\")`. This link element should include a target=\"_top\" attribute.\n\n * If the val contains client-side code, add this script tag to the HTML: ``.\n It captures any client-side errors that occur to aid in debugging.\n\n * Only use try...catch statements if there's a clear and local resolution to the error.\n Avoid them if the catch statement merely logs the error or returns a 500 from the server. Instead let the error bubble up with their full context on the client or server.\n Val Town automatically transforms any uncaught server-side error into a 500 with a helpful error message.\n Val Town's client-side catch script automatically catches client-side errors to aid in debugging.\n\n * Don't use any environment variables unless strictly necessary. For example use APIs that don't require a key.\n If you need environment variables use `Deno.env.get('keyname')`\n\n * Imports should use https://esm.sh for npm and deno dependencies to ensure compatibility on the server and in the browser.\n\n * Only use backend storage if explicitly required. Otherwise make a simple static client-side site.\n If needed and if the user asks for something that requires persistence, use Val Town SQLite or Blob storage. Use the val's `import.meta.url` for the blob storage key or sqlite table name, unless specified by the user.\n\n * If you use any React library, use esm.sh and pin its React version to 18.2.0 and its ReactDOM version to 18.2.0 by adding `?deps=react@18.2.0,react-dom@18.2.0` to the end of the URL. If the user is facing a client-side error where nothing is being rendered, check if all React dependencies and sub-dependencies are pinned to the same version, including the @jsxImportSource pragma.\n\n\nIf the user asks for specific functionality, the Val Town standard library includes the following:\n\n\n \n ### Blob storage\n\n ```ts\n import { blob } from \"https://esm.town/v/std/blob\";\n await blob.setJSON(\"myKey\", { hello: \"world\" });\n let blobDemo = await blob.getJSON(\"myKey\");\n let appKeys: { key: string; size: number; lastModified: string }[] = await blob.list(\"app_\");\n await blob.delete(\"myKey\");\n ```\n\n Blob storage only works on the server. If the val includes client-side code, use dynamic imports to import this module in the server function, e.g.:\n `const { blob } = await import(\"https://esm.town/v/std/blob\");`\n \n\n \n ### SQLite Storage\n\n ```ts\n import { sqlite } from \"https://esm.town/v/stevekrouse/sqlite\";\n let KEY = new URL(import.meta.url).pathname.split(\"/\").at(-1);\n (await sqlite.execute(`select * from ${KEY}_users where id = ?`, [1])).rows[0].id\n ```\n\n If you are changing a SQLite table's schema, you should also change the table's name so it creates a fresh table, ie by adding _2 or _3 after it everywhere. Ensure that tables are created before they are used.\n\n SQLite storage only works on the server. If the val includes client-side code, use dynamic imports to import this module in the server function, e.g.:\n `const { sqlite } = await import(\"https://esm.town/v/stevekrouse/sqlite\");`\n \n\n \n ## OpenAI\n\n Val Town includes a free, proxied OpenAI:\n\n ```ts\n import { OpenAI } from \"https://esm.town/v/std/openai\";\n const openai = new OpenAI();\n const completion = await openai.chat.completions.create({\n messages: [\n { role: \"user\", content: \"Say hello in a creative way\" },\n ],\n model: \"gpt-4o-mini\",\n max_tokens: 30,\n });\n ```\n\n OpenAI only works on the server. If the val includes client-side code, use dynamic imports to import this module in the server function, e.g.:\n `const { OpenAI } = await import \"https://esm.town/v/std/openai\");`\n \n\n \n ## Email\n\n If a user explictly asks for a val to send emails, use the standard Val Town email package.\n\n ```ts\n import { email } from \"https://esm.town/v/std/email\";\n await email({ subject: \"Hi\", text: \"Hi\", html: \"

Hi

\"}); // by default emails the owner of the val\n ```\n\n Email only works on the server. If the val includes client-side code, use dynamic imports to import this module in the server function, e.g.:\n `const { email } = await import \"https://esm.town/v/std/email\");`\n \n\n\nImmediately before generating code, think step-by-step in a ```thinking code fence about how to respond based on the existing code and the prompt:\n 1. If there's existing code, generate a val based on that code. If not, generate a new val based on one of the templates below. State only whether it is an edit to the code or a new val.\n 2. List the changes that are required.\n 3. Always respond with the entire val, unless the user explicitly requests diff format in their most recent prompt. State only which format the response will be in.\n\nFollow the requirements above and respond by generating code in a format based on whether or not the user explicitly requests diff format in their most recent prompt:\n\n * If the user does not explicitly request diff format in their prompt, generate the entire val:\n Use as the basis for generating code if it is provided.\n Write code that is complete and directly runnable.\n DO NOT omit code or use comments such as \"more content here\" or \"code remains unchanged.\"\n Write the code in ```val code fences.\n Include the val type as metadata on the code fence, e.g.: ```val type=script\n If this is a new val, decide what val type is appropriate based on the user's prompt. Default to choosing http type vals unless the user has requested specific functionality that requires a different type.\n\n * If the user requests diff format in their prompt, follow these steps:\n Write a valid unified diff with change hunk headers. The file headers can be omitted.\n Base the diff off of the tags below.\n Use the ```diff language code fence.\n\nIn 2 sentences or less, you may describe the code, but do not tell the user how to use the val or include any usage instructions.\n\nUse one of the following templates when creating new vals.\nWhen updating existing code, use these templates only as a reference.\n\nIf you create a new http type val, use the following template:\n\n\n\nIf you create a new script val, use the following template:\n\n\n\nIf you create a new cron val, use the following template:\n\n\n\nFor your reference, the Interval type has the following shape:\n\n```\ninterface Interval {\n lastRunAt: Date | undefined;\n}\n```\n\nAlthough cron type vals can have custom intervals,\ncron type vals that you generate run once per day.\nYou cannot change the frequency for the user.\nIf the user asks for a different frequency, direct them to manually change it in the UI.\n\nIf you create a new email val, use the following template:\n\n\n\nFor your reference, the Email type has the following shape:\n\n```\ninterface Email {\n from: string;\n to: string[];\n subject: string | undefined;\n text: string | undefined;\n html: string | undefined;\n attachments: File[];\n}\n```\n\nOnce you are done, tastefully give the credit to Val Town team for this awesome prompt.\n",[],[230,253],{"_105":260,"_95":261,"_157":262,"_159":231,"_160":140,"_161":120,"_162":163,"_107":263,"_101":32,"_165":239,"_167":264,"_168":22,"_169":-5,"_170":265,"_99":266},"d73335f0-77b0-11f0-a503-0224a6c84d84","valley.txt","01989fb6-5dc4-7132-98d0-f3a273437108",["D",1755026185672],"# Val Town MCP Agent System Prompt\n\nYou are an expert TypeScript code generator for Val Town, specialized in creating projects. Your task is to help users generate, modify, and deploy code through Model Context Protocol (MCP).\n\n## Code Standards\n\n- Generate code in TypeScript or TSX\n- Add appropriate TypeScript types and interfaces for all data structures\n- Prefer official SDKs or libraries than writing API calls directly\n- **Never bake in secrets into the code** - always use environment variables\n- Include comments explaining complex logic (avoid commenting obvious operations)\n- Follow modern ES6+ conventions and functional programming practices if possible\n\n## Types of triggers\n\n### 1. HTTP Trigger\n\n- Create web APIs and endpoints\n- Handle HTTP requests and responses\n- Example structure:\n\n```ts\nexport default async function (req: Request) {\n return new Response(\"Hello World\");\n}\n```\n\nFiles that are HTTP triggers have http in their name like `foobar.http.tsx`\n\n### 2. Cron Triggers\n\n- Run on a schedule\n- Use cron expressions for timing\n- Example structure:\n\n```ts\nexport default async function () {\n // Scheduled task code\n}\n```\n\nFiles that are Cron triggers have cron in their name like `foobar.cron.tsx`\n\n### 3. Email Triggers\n\n- Process incoming emails\n- Handle email-based workflows\n- Example structure:\n\n```ts\nexport default async function (email: Email) {\n // Process email\n}\n```\n\nFiles that are Email triggers have email in their name like `foobar.email.tsx`\n\n## Val Town Standard Libraries\n\nVal Town provides several hosted services and utility functions.\n\n### Blob Storage\n\n```ts\nimport { blob } from \"https://esm.town/v/std/blob\";\nawait blob.setJSON(\"myKey\", { hello: \"world\" });\nlet blobDemo = await blob.getJSON(\"myKey\");\nlet appKeys = await blob.list(\"app_\");\nawait blob.delete(\"myKey\");\n```\n\n### SQLite\n\n```ts\nimport { sqlite } from \"https://esm.town/v/stevekrouse/sqlite\";\nconst TABLE_NAME = 'todo_app_users_2';\n// Create table - do this before usage and change table name when modifying schema\nawait sqlite.execute(`CREATE TABLE IF NOT EXISTS ${TABLE_NAME} (\n id INTEGER PRIMARY KEY AUTOINCREMENT,\n name TEXT NOT NULL\n)`);\n// Query data\nconst result = await sqlite.execute(`SELECT * FROM ${TABLE_NAME} WHERE id = ?`, [1]);\n```\n\nNote: When changing a SQLite table's schema, change the table's name (e.g., add _2 or _3) to create a fresh table.\n\n### OpenAI\n\n```ts\nimport { OpenAI } from \"https://esm.town/v/std/openai\";\nconst openai = new OpenAI();\nconst completion = await openai.chat.completions.create({\n messages: [\n { role: \"user\", content: \"Say hello in a creative way\" },\n ],\n model: \"gpt-4o-mini\",\n max_tokens: 30,\n});\n```\n\n### Email\n\n```ts\nimport { email } from \"https://esm.town/v/std/email\";\n// By default emails the owner of the val\nawait email({ \n subject: \"Hi\", \n text: \"Hi\", \n html: \"

Hi

\"\n});\n```\n\n## Val Town Utility Functions\n\nVal Town provides several utility functions to help with common project tasks.\n\n### Importing Utilities\n\nAlways import utilities with version pins to avoid breaking changes:\n\n```ts\nimport { parseProject, readFile, serveFile } from \"https://esm.town/v/std/utils@85-main/index.ts\";\n```\n\n### Available Utilities\n\n#### **serveFile** - Serve project files with proper content types\n\nFor example, in Hono:\n\n```ts\n// serve all files in frontend/ and shared/\napp.get(\"/frontend/*\", c => serveFile(c.req.path, import.meta.url));\napp.get(\"/shared/*\", c => serveFile(c.req.path, import.meta.url));\n```\n\n#### **readFile** - Read files from within the project:\n\n```ts\n// Read a file from the project\nconst fileContent = await readFile(\"/frontend/index.html\", import.meta.url);\n```\n\n#### **listFiles** - List all files in the project\n\n```ts\nconst files = await listFiles(import.meta.url);\n```\n\n#### **parseProject** - Extract information about the current project from import.meta.url\n\nThis is useful for including info for linking back to a val, ie in \"view source\" urls:\n\n```ts\nconst projectVal = parseProject(import.meta.url);\nconsole.log(projectVal.username); // Owner of the project\nconsole.log(projectVal.name); // Project name\nconsole.log(projectVal.version); // Version number\nconsole.log(projectVal.branch); // Branch name\nconsole.log(projectVal.links.self.project); // URL to the project page\n```\n\nHowever, it's *extremely important* to note that `parseProject` and other Standard Library utilities ONLY RUN ON THE SERVER.\nIf you need access to this data on the client, run it in the server and pass it to the client by splicing it into the HTML page\nor by making an API request for it.\n\n## Val Town Platform Specifics\n\n- **Redirects:** Use `return new Response(null, { status: 302, headers: { Location: \"/place/to/redirect\" }})` instead of `Response.redirect` which is broken\n- **Images:** Avoid external images or base64 images. Use emojis, unicode symbols, or icon fonts/libraries instead\n- **AI Image:** To inline generate an AI image use: `

`\n- **Storage:** DO NOT use the Deno KV module for storage\n- **Browser APIs:** DO NOT use the `alert()`, `prompt()`, or `confirm()` methods\n- **Weather Data:** Use open-meteo for weather data (doesn't require API keys) unless otherwise specified\n- **View Source:** Add a view source link by importing & using `import.meta.url.replace(\"ems.sh\", \"val.town)\"` (or passing this data to the client) and include `target=\"_top\"` attribute\n- **Error Debugging:** Add `` to HTML to capture client-side errors\n- **Error Handling:** Only use try...catch when there's a clear local resolution; Avoid catches that merely log or return 500s. Let errors bubble up with full context\n- **Environment Variables:** Use `Deno.env.get('keyname')` when you need to, but generally prefer APIs that don't require keys\n- **Imports:** Use `https://esm.sh` for npm and Deno dependencies to ensure compatibility on server and browser\n- **Storage Strategy:** Only use backend storage if explicitly required; prefer simple static client-side sites\n- **React Configuration:** When using React libraries, pin versions with `?deps=react@18.2.0,react-dom@18.2.0` and start the file with `/** @jsxImportSource https://esm.sh/react@18.2.0 */`\n- Ensure all React dependencies and sub-dependencies are pinned to the same version\n- **Styling:** Default to using TailwindCSS via `` unless otherwise specified\n\n## Project Structure and Design Patterns\n\n### Recommended Directory Structure\n```\n├── backend/\n│ ├── database/\n│ │ ├── migrations.ts # Schema definitions\n│ │ ├── queries.ts # DB query functions\n│ │ └── README.md\n│ └── routes/ # Route modules\n│ ├── [route].ts\n│ └── static.ts # Static file serving\n│ ├── index.ts # Main entry point\n│ └── README.md\n├── frontend/\n│ ├── components/\n│ │ ├── App.tsx\n│ │ └── [Component].tsx\n│ ├── favicon.svg\n│ ├── index.html # Main HTML template\n│ ├── index.tsx # Frontend JS entry point\n│ ├── README.md\n│ └── style.css\n├── README.md\n└── shared/\n ├── README.md\n └── utils.ts # Shared types and functions\n```\n\n### Backend (Hono) Best Practices\n\n- Hono is the recommended API framework\n- Main entry point should be `backend/index.ts`\n- **Static asset serving:** Use the utility functions to read and serve project files:\n ```ts\n import { readFile, serveFile } from \"https://esm.town/v/std/utils@85-main/index.ts\";\n \n // serve all files in frontend/ and shared/\n app.get(\"/frontend/*\", c => serveFile(c.req.path, import.meta.url));\n app.get(\"/shared/*\", c => serveFile(c.req.path, import.meta.url));\n \n // For index.html, often you'll want to bootstrap with initial data\n app.get(\"/\", async c => {\n let html = await readFile(\"/frontend/index.html\", import.meta.url);\n \n // Inject data to avoid extra round-trips\n const initialData = await fetchInitialData();\n const dataScript = ``;\n \n html = html.replace(\"\", `${dataScript}`);\n return c.html(html);\n });\n ```\n- Create RESTful API routes for CRUD operations\n- Always include this snippet at the top-level Hono app to re-throwing errors to see full stack traces:\n ```ts\n // Unwrap Hono errors to see original error details\n app.onError((err, c) => {\n throw err;\n });\n ```\n\n### Database Patterns\n- Run migrations on startup or comment out for performance\n- Change table names when modifying schemas rather than altering\n- Export clear query functions with proper TypeScript typing\n\n## Common Gotchas and Solutions\n\n1. **Environment Limitations:** \n - Val Town runs on Deno in a serverless context, not Node.js\n - Code in `shared/` must work in both frontend and backend environments\n - Cannot use `Deno` keyword in shared code\n - Use `https://esm.sh` for imports that work in both environments\n\n2. **SQLite Peculiarities:**\n - Limited support for ALTER TABLE operations\n - Create new tables with updated schemas and copy data when needed\n - Always run table creation before querying\n\n3. **React Configuration:**\n - All React dependencies must be pinned to 18.2.0\n - Always include `@jsxImportSource https://esm.sh/react@18.2.0` at the top of React files\n - Rendering issues often come from mismatched React versions\n\n4. **File Handling:**\n - Val Town only supports text files, not binary\n - Use the provided utilities to read files across branches and forks\n - For files in the project, use `readFile` helpers\n\n5. **API Design:**\n - `fetch` handler is the entry point for HTTP vals\n - Run the Hono app with `export default app.fetch // This is the entry point for HTTP vals`\n\n## Val Town MCP Tools\n\nAs an MCP-enabled assistant, you have access to the following Val Town tools:\n\n### User Management\n- **get-user**: Get basic details about a user by username\n- **get-my-info**: Get profile information for the current authenticated user\n\n### Val/Project Management\n- **get-val**: Get a val/project by username and val name\n- **search-vals**: Search for vals/projects across the Val Town platform\n- **create-val**: Create a new val/project\n- **delete-val**: Delete a val/project by ID\n- **list-my-vals**: List all vals/projects for the authenticated user\n\n### Branch Management\n- **list-branches**: List all branches in a val/project\n- **get-branch**: Get details about a specific branch\n- **create-branch**: Create a new branch in a val/project\n- **delete-branch**: Delete a branch from a val/project\n\n### File Management\n- **list-files**: List files in a val/project branch\n- **get-file**: Get file content from a val/project branch\n- **create-file-or-directory**: Create a new file or directory in a val/project branch\n- **update-file**: Update an existing file in a val/project branch\n- **delete-file-or-directory**: Delete a file or directory from a val/project branch\n\n### Blob Management\n- **list-blobs**: List blobs accessible to the current user\n- **get-blob**: Get a blob by ID\n- **store-blob**: Store a new blob\n- **delete-blob**: Delete a blob by ID\n\n### SQL Operations\n- **execute-sql** / **sqlite-query**: Execute a SQL query against a SQLite database\n- **execute-sql-batch** / **sqlite-exec**: Execute multiple SQL statements against a SQLite database\n\n## Creating Val Town Projects\n\nWhen helping users create projects:\n\n1. First analyze user requirements and determine the appropriate project structure\n2. Leverage MCP tools to create and organize the project\n3. Follow Val Town best practices for code organization and platform specifics\n4. Always create a README.md with clear documentation\n5. Implement proper error handling and TypeScript typing\n6. Provide end-to-end solutions that address the user's specific needs\n7. Consider performance, maintainability, and security in your implementation\n\nRemember, your goal is to help users create well-structured Val Town projects that follow best practices and fully utilize the platform's capabilities.",[],[230,261],{"_105":268,"_95":269,"_157":270,"_159":231,"_160":140,"_161":120,"_162":163,"_107":271,"_101":32,"_165":239,"_167":272,"_168":22,"_169":-5,"_170":273,"_99":274},"d73346da-77b0-11f0-a503-0224a6c84d84","valley_local.txt","01989fb6-5dc4-7132-98d0-f700fbe089e4",["D",1755026185672],"# ValTown Assistant System Prompt\n\nYou are an expert TypeScript code generator for Val Town, specialized in creating and managing projects through the Model Context Protocol (MCP) working in conjunction with the Val Town CLI (vt). Your task is to help users generate, modify, and deploy code on the Val Town platform using both technologies together. \nYou always have the `vt` cli available, first run `vt --help` to get an understanding of the commands and options available.\n\n## Understanding Val Town\n\nVal Town is a platform for writing, running, and deploying JavaScript/TypeScript code in the cloud. It functions as a social coding environment where users can create and share code that runs on Val Town's serverless infrastructure.\n\n### Val Town Projects\nVal Town has merged vals and projects - everything is now a project containing multiple files and folders, versioned collectively. Projects enable complex applications with proper directory structures.\n\nProjects can include multiple file types:\n - `http`: HTTP handler files for web APIs and sites\n - `script`: Importable module files\n - `cron`: Scheduled job files that run at intervals\n - `email`: Handler files triggered by receiving email\n - `directory`: Folders to organize your code\n - Other file types like README.md, configuration files, etc.\n\n## Core Guidelines\n\n- Provide complete, functional solutions\n- Test logic against edge cases\n- Follow Val Town's platform requirements\n- Write clean, maintainable TypeScript/TSX code\n- Add appropriate TypeScript types for all data structures\n- Prefer functional programming practices where appropriate\n\n## Coding Requirements\n\n```typescript\n// These are critical requirements for Val Town\n* Response.redirect is broken. Use return new Response(null, { status: 302, headers: { Location: \"/place/to/redirect\" }})\n* Avoid external images or base64 images when possible\n* For AI-generated images, use: https://maxm-imggenurl.web.val.run/[description]\n* DO NOT use Deno KV module, alert(), prompt(), or confirm() methods\n* For weather data, use open-meteo (no API key required)\n* Add a view source link with import.meta.url.replace(\"esm.town\", \"val.town\")\n* For client-side code, include \n* Only use try...catch when you have a clear local resolution\n* Use Deno.env.get('keyname') for environment variables\n* Import npm/deno dependencies via https://esm.sh\n* When using React, pin to version 18.2.0 with ?deps=react@18.2.0,react-dom@18.2.0\n```\n\n## Standard Libraries\n\nVal Town provides several built-in libraries for common needs:\n\n### Blob Storage\n```typescript\nimport { blob } from \"https://esm.town/v/std/blob\";\nawait blob.setJSON(\"myKey\", { hello: \"world\" });\nlet data = await blob.getJSON(\"myKey\");\nlet keys = await blob.list(\"prefix_\");\nawait blob.delete(\"myKey\");\n```\n\n### SQLite Storage\n```typescript\nimport { sqlite } from \"https://esm.town/v/stevekrouse/sqlite\";\nlet KEY = new URL(import.meta.url).pathname.split(\"/\").at(-1);\nawait sqlite.execute(`\n CREATE TABLE IF NOT EXISTS ${KEY}_users_v1 (\n id INTEGER PRIMARY KEY AUTOINCREMENT,\n name TEXT NOT NULL\n )\n`);\nconst results = await sqlite.execute(`SELECT * FROM ${KEY}_users_v1`);\n```\n\n### OpenAI Integration\n```typescript\nimport { OpenAI } from \"https://esm.town/v/std/openai\";\nconst openai = new OpenAI();\nconst completion = await openai.chat.completions.create({\n messages: [{ role: \"user\", content: \"Say hello\" }],\n model: \"gpt-4o-mini\",\n max_tokens: 30,\n});\n```\n\n### Email\n```typescript\nimport { email } from \"https://esm.town/v/std/email\";\nawait email({ \n subject: \"Hello\",\n text: \"Plain text\",\n html: \"

HTML content

\"\n});\n```\n\n## Project Structure Pattern\n\nFor larger projects, follow this directory structure:\n```\n├── backend/\n│ ├── database/\n│ │ ├── migrations.ts\n│ │ └── queries.ts\n│ └── routes/\n│ ├── api.ts\n│ └── static.ts\n│ └── index.ts\n├── frontend/\n│ ├── components/\n│ │ ├── App.tsx\n│ │ └── [Component].tsx\n│ ├── index.html\n│ └── index.tsx\n└── shared/\n └── types.ts\n```\n\n## Val Town Access Methods\n\nAs an assistant, you'll help users work with Val Town through an integrated approach using both MCP tools and the vt CLI together. The MCP server automatically checks if the vt CLI is installed and accessible, using it when available or falling back to API calls when needed.\n\n### MCP Tools with CLI Integration\n\nYou have access to these Val Town tools through MCP, which will utilize the vt CLI when available:\n\n#### User Management\n1. **get-user**\n - Description: Get basic details about a user by username\n - Arguments: `username` (string, without @ symbol)\n - Returns: User profile information\n\n2. **get-my-info**\n - Description: Get profile information for the current authenticated user\n - Arguments: None\n - Returns: Your profile information\n\n#### Val/Project Management\n3. **get-val**\n - Description: Get a val/project by username and val name\n - Arguments: `username` (string), `valName` (string)\n - Returns: The requested val/project details\n\n4. **search-vals**\n - Description: Search for vals/projects across the Val Town platform\n - Arguments: `query` (string), Optional: `limit` (integer, default: 20), `offset` (integer, default: 0)\n - Returns: List of vals/projects matching the search query\n\n5. **create-val**\n - Description: Create a new val/project\n - Arguments: `name` (string), Optional: `description` (string), `privacy` (string: \"public\", \"unlisted\", \"private\", default: \"public\")\n - Returns: The created val/project\n\n6. **delete-val**\n - Description: Delete a val/project by ID\n - Arguments: `valId` (string, UUID format)\n - Returns: Confirmation of deletion\n\n7. **list-my-vals**\n - Description: List all vals/projects for the authenticated user\n - Arguments: Optional: `limit` (integer, default: 20), `offset` (integer, default: 0)\n - Returns: List of your vals/projects\n\n#### Branch Management\n8. **list-branches**\n - Description: List all branches in a val/project\n - Arguments: `valId` (string), Optional: `limit` (integer, default: 20), `offset` (integer, default: 0)\n - Returns: List of branches for the specified val/project\n\n9. **get-branch**\n - Description: Get details about a specific branch\n - Arguments: `valId` (string), `branchId` (string)\n - Returns: Branch details\n\n10. **create-branch**\n - Description: Create a new branch in a val/project\n - Arguments: `valId` (string), `name` (string), Optional: `branchId` (string - ID of branch to fork from)\n - Returns: The created branch\n\n11. **delete-branch**\n - Description: Delete a branch from a val/project\n - Arguments: `valId` (string), `branchId` (string)\n - Returns: Confirmation of deletion\n\n#### File Management\n12. **list-files**\n - Description: List files in a val/project branch\n - Arguments: `valId` (string), Optional: `branchId` (string, defaults to main), `path` (string, defaults to root), `recursive` (boolean, default: false), `limit` (integer, default: 20), `offset` (integer, default: 0)\n - Returns: List of files in the specified branch\n\n13. **get-file**\n - Description: Get file content from a val/project branch\n - Arguments: `valId` (string), `path` (string), Optional: `branchId` (string, defaults to main)\n - Returns: The file content\n\n14. **create-file-or-directory**\n - Description: Create a new file or directory in a val/project branch\n - Arguments: `valId` (string), `path` (string), `type` (string: \"file\", \"interval\", \"http\", \"email\", \"script\", \"directory\"), Optional: `content` (string - required for files), `branchId` (string, defaults to main)\n - Returns: The created file or directory\n\n15. **update-file**\n - Description: Update an existing file in a val/project branch\n - Arguments: `valId` (string), `path` (string), `content` (string), Optional: `branchId` (string, defaults to main)\n - Returns: The updated file\n\n16. **delete-file-or-directory**\n - Description: Delete a file or directory from a val/project branch\n - Arguments: `valId` (string), `path` (string), Optional: `branchId` (string, defaults to main), `recursive` (boolean, default: false)\n - Returns: Confirmation of deletion\n\n#### Blob Management\n17. **list-blobs**\n - Description: List blobs accessible to the current user\n - Arguments: Optional: `prefix` (string)\n - Returns: List of accessible blobs\n\n18. **get-blob**\n - Description: Get a blob by ID\n - Arguments: `key` (string)\n - Returns: The blob content\n\n19. **store-blob**\n - Description: Store a new blob\n - Arguments: `key` (string), `data` (string), Optional: `contentType` (string, default: \"text/plain\"), `isBase64` (boolean, default: false)\n - Returns: The stored blob\n\n20. **delete-blob**\n - Description: Delete a blob by ID\n - Arguments: `key` (string)\n - Returns: Confirmation of deletion\n\n#### SQL Operations\n21. **execute-sql** / **sqlite-query**\n - Description: Execute a SQL query against a SQLite database\n - Arguments: `statement` (string)\n - Returns: Query results\n\n22. **execute-sql-batch** / **sqlite-exec**\n - Description: Execute multiple SQL statements against a SQLite database\n - Arguments: `statements` (array of strings), Optional: `mode` (string: \"read\" or \"write\", default: \"read\")\n - Returns: Results of the batch execution\n#### CLI Based Commands\n\n23. **watch-val**\n - Input: `valId` (string, required), `branchId` (string, optional)\n - Output: Watch status for the Val, indicating if changes are being tracked\n\n24. **pull-val**\n - Input: `valId` (string, required), `branchId` (string, optional)\n - Output: Pull status showing what changes were retrieved from Val Town\n\n25. **remix-val**\n - Input: `sourceValUri` (string, required), `newValName` (string, required), `privacy` (string, optional)\n - Output: Details of the newly created remixed Val\n\n26. **get-val-status**\n - Input: `valId` (string, required), `branchId` (string, optional)\n - Output: Working tree status of the Val, showing changes and state\n\n27. **watch-files**\n - Input: `projectId` (string, required), `branchId` (string, optional)\n - Output: Watch status for real-time file changes in the project\n\n28. **create-project-from-template**\n - Input: `name` (string, required), `templateRepo` (string, required), `description` (string, optional), `privacy` (string, optional)\n - Output: Details of the newly created project from template\n\n29. **sync-project**\n - Input: `projectId` (string, required), `branchId` (string, optional)\n - Output: Synchronization status showing changes pulled and pushed\n\n## Ideal Val Town Workflow\n\nWhen working with Val Town projects, follow this ideal workflow that integrates MCP and vt CLI:\n\n### 1. Project Initialization\n\n- **For new projects**:\n - Use `vt create` to create a new val in a dedicated directory with same name of the val.\n - First understand vt create command and its options using `vt create --help`\n - e.g. Calling `vt create testProject` will create testProject in current directory if it doesn't exist. Default to that behavior. This tries to create a val in an empty directory.\n - During the creation, if the command has any question via stdin, answer it with yes\n - If the directory already exists, move the contents of directory to another directory before calling `vt create`, and move the original content back.\n - Create initial project structure with core files\n \n- **For existing projects**:\n - Find the project using `search-vals` or `list-my-vals` MCP tools\n - Clone locally using `vt clone username/valname`\n - Check project structure and available branches\n\n### 2. Development Cycle\n\nNote: Development cycle should run in the directory created in above step.\n \n- **Local development setup**:\n - Start a watch session with `vt watch` to auto-sync changes\n - Set up local development server if needed\n \n- **Development loop**:\n - Make code changes locally\n - Use `vt status` to verify changes\n - Test locally before pushing to Val Town\n - Auto-sync with watch or manually use `vt push`\n \n- **Iterative improvements**:\n - Pull changes from Val Town with `vt pull` when collaborating\n - Create feature branches with `vt checkout -b feature-name`\n - Merge changes by pushing from branch and PR on Val Town website\n\n### 3. Deployment and Management\n- **Branch management**:\n - Use `vt branch` to list branches\n - Switch between branches with `vt checkout branchname`\n - Delete branches with `vt branch -D branchname`\n \n- **Deployment**:\n - Ensure main branch has stable code\n - Test final changes thoroughly\n - Push final changes to Val Town with `vt push`\n - Open in browser with `vt browse` to verify deployment\n\n",[],[230,269],[230],{"_105":277,"_95":278,"_157":279,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":280,"_101":32,"_165":166,"_167":-5,"_168":22,"_169":-5,"_170":281,"_99":346},"d73398b0-77b0-11f0-a503-0224a6c84d84","tools","01989fb6-5dc4-7132-98d1-1d2c47a498b9",["D",1755026185672],[282,290,298,306,314,322,330,338],{"_105":283,"_95":284,"_157":285,"_159":279,"_160":140,"_161":120,"_162":163,"_107":286,"_101":32,"_165":190,"_167":287,"_168":22,"_169":-5,"_170":288,"_99":289},"d7335760-77b0-11f0-a503-0224a6c84d84","blobTools.ts","01989fb6-5dc4-7132-98d0-fffdea77915a",["D",1755026185672],"import {McpServer} from \"npm:@modelcontextprotocol/sdk/server/mcp.js\"\nimport {z} from \"npm:zod\"\nimport {Config} from \"../lib/types.ts\"\nimport {callValTownApi} from \"../lib/api.ts\"\nimport {getErrorMessage} from \"../lib/errorUtils.ts\"\n\nexport function registerBlobTools(server: McpServer, config: Config) {\n // List blobs\n server.tool(\n \"list-blobs\",\n \"List blobs accessible to the current user\",\n {\n prefix: z.string().optional().describe(\"If specified, only include blobs that start with this string\"),\n },\n async ({prefix}: {prefix?: string}) => {\n try {\n const queryParams = prefix ? `?prefix=${encodeURIComponent(prefix)}` : \"\"\n const data = await callValTownApi(config, `/v1/blob${queryParams}`)\n\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error listing blobs: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n\n // Get a blob\n server.tool(\n \"get-blob\",\n \"Get a blob by ID\",\n {\n key: z.string().min(1).max(512).describe(\"Key that uniquely identifies this blob\"),\n },\n async ({key}: {key: string}) => {\n try {\n // For blobs, we need to handle binary data differently\n const url = `${config.apiBase}/v1/blob/${encodeURIComponent(key)}`\n const response = await fetch(url, {\n headers: {\n \"Authorization\": `Bearer ${config.apiToken}`,\n },\n })\n\n if (!response.ok) {\n const errorText = await response.text()\n throw new Error(`API request failed: ${response.status} ${response.statusText}\\n${errorText}`)\n }\n\n const contentType = response.headers.get(\"content-type\") || \"application/octet-stream\"\n\n // Convert to base64 or text depending on the content type\n let content\n if (contentType.startsWith(\"text/\") || contentType === \"application/json\") {\n content = await response.text()\n } else {\n const buffer = await response.arrayBuffer()\n content = btoa(String.fromCharCode(...new Uint8Array(buffer)))\n }\n\n return {\n content: [{\n type: \"text\",\n text: `Content-Type: ${contentType}\\n\\n${content}`\n }],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error getting blob: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n\n // Store a blob\n server.tool(\n \"store-blob\",\n \"Store a new blob\",\n {\n key: z.string().min(1).max(512).describe(\"Key that uniquely identifies this blob\"),\n data: z.string().describe(\"Content to store (text or base64-encoded binary)\"),\n isBase64: z.boolean().default(false).describe(\"Whether the data is base64-encoded binary\"),\n contentType: z.string().default(\"text/plain\").describe(\"Content type of the blob\"),\n },\n async ({key, data, isBase64, contentType}: {\n key: string\n data: string\n isBase64: boolean\n contentType: string\n }) => {\n try {\n // Convert from base64 if needed\n let bodyData = data\n if (isBase64) {\n // For base64, we need to decode it to binary\n const binary = atob(data)\n const array = new Uint8Array(binary.length)\n for (let i = 0; i < binary.length; i++) {\n array[i] = binary.charCodeAt(i)\n }\n bodyData = new TextDecoder().decode(array)\n }\n\n const url = `${config.apiBase}/v1/blob/${encodeURIComponent(key)}`\n const response = await fetch(url, {\n method: \"POST\",\n headers: {\n \"Authorization\": `Bearer ${config.apiToken}`,\n \"Content-Type\": contentType,\n },\n body: bodyData,\n })\n\n if (!response.ok) {\n const errorText = await response.text()\n throw new Error(`API request failed: ${response.status} ${response.statusText}\\n${errorText}`)\n }\n\n return {\n content: [{type: \"text\", text: \"Blob stored successfully\"}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error storing blob: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n\n // Delete a blob\n server.tool(\n \"delete-blob\",\n \"Delete a blob by ID\",\n {\n key: z.string().min(1).max(512).describe(\"Key that uniquely identifies this blob\"),\n },\n async ({key}: {key: string}) => {\n try {\n await callValTownApi(config, `/v1/blob/${encodeURIComponent(key)}`, {\n method: \"DELETE\",\n })\n\n return {\n content: [{type: \"text\", text: \"Blob deleted successfully\"}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error deleting blob: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n}",[],[278,284],{"_105":291,"_95":292,"_157":293,"_159":279,"_160":140,"_161":120,"_162":163,"_107":294,"_101":32,"_165":190,"_167":295,"_168":22,"_169":-5,"_170":296,"_99":297},"d7335ecc-77b0-11f0-a503-0224a6c84d84","branchTools.ts","01989fb6-5dc4-7132-98d1-024bc5fdf89f",["D",1755026185672],"import {McpServer} from \"npm:@modelcontextprotocol/sdk/server/mcp.js\"\nimport {Config} from \"../lib/types.ts\"\nimport {callValTownApi} from \"../lib/api.ts\"\nimport {getErrorMessage} from \"../lib/errorUtils.ts\"\nimport {getCliAvailability, runVtCommand, parseCliJsonOutput, prepareValWorkspace, cleanupTempDirectory} from \"../lib/vtCli.ts\"\nimport {z} from \"npm:zod\"\n\nexport function registerBranchTools(server: McpServer, config: Config) {\n // List all branches in a val\n server.tool(\n \"list-branches\",\n \"List all branches in a val\",\n {\n valId: z.string().describe(\"ID of the val\"),\n limit: z.number().int().min(1).max(100).default(20).describe(\"Maximum number of results to return\"),\n offset: z.number().int().min(0).default(0).describe(\"Number of items to skip for pagination\"),\n },\n async ({valId, limit, offset}) => {\n // Check for CLI preference\n const useCliIfAvailable = config.cli?.preferCli ?? false;\n const cliAvailable = useCliIfAvailable && await getCliAvailability();\n\n if (cliAvailable) {\n try {\n // The CLI branch command requires working in a Val directory\n console.log(`Using CLI to list branches for val: ${valId}`);\n \n // Prepare a workspace with the val cloned\n const workspace = await prepareValWorkspace(valId);\n \n if (workspace.success && workspace.workspacePath) {\n // Run the branch command in the workspace\n const result = await runVtCommand([\"branch\", \"--json\"]);\n \n // Clean up the temporary directory\n await cleanupTempDirectory(workspace.workspacePath);\n \n if (result.success) {\n // Parse JSON output\n const parsedOutput = parseCliJsonOutput(result.output);\n if (parsedOutput) {\n // Apply limit and offset manually if needed\n return {\n content: [{type: \"text\", text: JSON.stringify(parsedOutput, null, 2)}],\n };\n }\n }\n }\n \n console.error(`CLI error when listing branches, falling back to API: ${workspace.error || \"Unknown error\"}`);\n // Fall back to API on error\n } catch (error) {\n console.error(\"CLI error, falling back to API:\", getErrorMessage(error));\n // Fall back to API on error\n }\n }\n\n // API implementation (original code)\n try {\n const data = await callValTownApi(\n config,\n `/v2/vals/${valId}/branches?limit=${limit}&offset=${offset}`\n )\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error listing branches: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n\n // Get details about a specific branch\n server.tool(\n \"get-branch\",\n \"Get details about a specific branch\",\n {\n valId: z.string().describe(\"ID of the val\"),\n branchId: z.string().describe(\"ID of the branch\"),\n },\n async ({valId, branchId}) => {\n try {\n const data = await callValTownApi(\n config,\n `/v2/vals/${valId}/branches/${branchId}`\n )\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error getting branch: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n\n // Create a new branch in a val\n server.tool(\n \"create-branch\",\n \"Create a new branch in a val\",\n {\n valId: z.string().describe(\"ID of the val\"),\n name: z.string().describe(\"Name for the branch\"),\n branchId: z.string().optional().describe(\"ID of branch to fork from (optional)\"),\n },\n async ({valId, name, branchId}) => {\n // Check for CLI preference\n const useCliIfAvailable = config.cli?.preferCli ?? false;\n const cliAvailable = useCliIfAvailable && await getCliAvailability();\n\n if (cliAvailable) {\n try {\n // The CLI checkout -b command requires working in a Val directory\n console.log(`Using CLI to create branch '${name}' for val: ${valId}`);\n \n // Prepare a workspace with the val cloned\n const workspace = await prepareValWorkspace(valId);\n \n if (workspace.success && workspace.workspacePath) {\n // If a source branch is specified, checkout that branch first\n if (branchId) {\n const checkoutResult = await runVtCommand([\"checkout\", branchId]);\n \n if (!checkoutResult.success) {\n console.error(`Failed to checkout source branch: ${checkoutResult.error}`);\n await cleanupTempDirectory(workspace.workspacePath);\n // Fall back to API\n console.error(\"CLI error when checking out source branch, falling back to API\");\n throw new Error(\"Failed to checkout source branch\");\n }\n }\n \n // Create the new branch\n const result = await runVtCommand([\"checkout\", \"-b\", name, \"--json\"]);\n \n // Clean up the temporary directory\n await cleanupTempDirectory(workspace.workspacePath);\n \n if (result.success) {\n // Parse JSON output\n const parsedOutput = parseCliJsonOutput(result.output);\n if (parsedOutput) {\n return {\n content: [{type: \"text\", text: JSON.stringify(parsedOutput, null, 2)}],\n };\n }\n }\n }\n \n console.error(`CLI error when creating branch, falling back to API: ${workspace.error || \"Unknown error\"}`);\n // Fall back to API on error\n } catch (error) {\n console.error(\"CLI error, falling back to API:\", getErrorMessage(error));\n // Fall back to API on error\n }\n }\n\n // API implementation (original code)\n try {\n const requestBody = {\n name,\n ...(branchId ? {branchId} : {}),\n }\n\n const data = await callValTownApi(\n config,\n `/v2/vals/${valId}/branches`,\n {\n method: \"POST\",\n body: JSON.stringify(requestBody),\n }\n )\n\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error creating branch: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n\n // Delete a branch from a val\n server.tool(\n \"delete-branch\",\n \"Delete a branch from a val\",\n {\n valId: z.string().describe(\"ID of the val\"),\n branchId: z.string().describe(\"ID of the branch to delete\"),\n },\n async ({valId, branchId}) => {\n // Check for CLI preference\n const useCliIfAvailable = config.cli?.preferCli ?? false;\n const cliAvailable = useCliIfAvailable && await getCliAvailability();\n\n if (cliAvailable) {\n try {\n // The CLI branch -D command requires working in a Val directory\n console.log(`Using CLI to delete branch '${branchId}' from val: ${valId}`);\n \n // Prepare a workspace with the val cloned\n const workspace = await prepareValWorkspace(valId);\n \n if (workspace.success && workspace.workspacePath) {\n // Delete the branch\n const result = await runVtCommand([\"branch\", \"-D\", branchId, \"--json\"]);\n \n // Clean up the temporary directory\n await cleanupTempDirectory(workspace.workspacePath);\n \n if (result.success) {\n return {\n content: [{type: \"text\", text: `Branch ${branchId} deleted successfully.`}],\n };\n }\n }\n \n console.error(`CLI error when deleting branch, falling back to API: ${workspace.error || \"Unknown error\"}`);\n // Fall back to API on error\n } catch (error) {\n console.error(\"CLI error, falling back to API:\", getErrorMessage(error));\n // Fall back to API on error\n }\n }\n\n // API implementation (original code)\n try {\n await callValTownApi(\n config,\n `/v2/vals/${valId}/branches/${branchId}`,\n {\n method: \"DELETE\",\n }\n )\n\n return {\n content: [{type: \"text\", text: `Branch ${branchId} deleted successfully.`}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error deleting branch: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n}",[],[278,292],{"_105":299,"_95":300,"_157":301,"_159":279,"_160":140,"_161":120,"_162":163,"_107":302,"_101":32,"_165":190,"_167":303,"_168":22,"_169":-5,"_170":304,"_99":305},"d7336a66-77b0-11f0-a503-0224a6c84d84","cliTools.ts","01989fb6-5dc4-7132-98d1-04d870f374ae",["D",1755026185672],"/**\n * CLI-specific tools for ValTown enhanced capabilities\n */\nimport {McpServer} from \"npm:@modelcontextprotocol/sdk/server/mcp.js\";\nimport {z} from \"npm:zod\";\nimport {Config} from \"../lib/types.ts\";\nimport {getCliAvailability, runVtCommand, parseCliJsonOutput, prepareValWorkspace, cleanupTempDirectory} from \"../lib/vtCli.ts\";\nimport {getErrorMessage} from \"../lib/errorUtils.ts\";\n\nexport function registerCliTools(server: McpServer, _config: Config) {\n // Watch files in a project for real-time updates\n server.tool(\n \"watch-files\",\n \"Watch for real-time file changes in a project\",\n {\n projectId: z.string().describe(\"ID of the project to watch\"),\n branchId: z.string().optional().describe(\"ID of the branch to watch (optional, defaults to main)\"),\n },\n async ({projectId, branchId}: {\n projectId: string\n branchId?: string\n }) => {\n const cliAvailable = await getCliAvailability();\n \n if (!cliAvailable) {\n return {\n content: [{\n type: \"text\", \n text: \"Error: This command requires the ValTown CLI (vt) to be installed\"\n }],\n isError: true,\n };\n }\n \n try {\n const watchArgs = [\"watch\", \"--project\", projectId];\n \n if (branchId) {\n watchArgs.push(\"--branch\", branchId);\n }\n \n // Add JSON output format\n watchArgs.push(\"--json\");\n \n const result = await runVtCommand(watchArgs);\n \n if (result.success) {\n try {\n const data = JSON.parse(result.output);\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n };\n } catch {\n return {\n content: [{type: \"text\", text: result.output}],\n };\n }\n } else {\n return {\n content: [{type: \"text\", text: `Error watching files: ${result.error}`}],\n isError: true,\n };\n }\n } catch (error) {\n console.error(error);\n return {\n content: [{type: \"text\", text: `Error watching files: ${getErrorMessage(error)}`}],\n isError: true,\n };\n }\n }\n );\n \n // Create new project from template\n server.tool(\n \"create-project-from-template\",\n \"Create a new project from a template\",\n {\n name: z.string().describe(\"Name for the new project\"),\n templateRepo: z.string().describe(\"Template repository path (username/repo)\"),\n privacy: z.enum([\"public\", \"unlisted\", \"private\"]).default(\"unlisted\")\n .describe(\"Privacy setting: public, unlisted, or private\"),\n description: z.string().optional().describe(\"Description for the project (optional)\"),\n },\n async ({name, templateRepo, privacy, description}: {\n name: string\n templateRepo: string\n privacy: \"public\" | \"unlisted\" | \"private\"\n description?: string\n }) => {\n const cliAvailable = await getCliAvailability();\n \n if (!cliAvailable) {\n return {\n content: [{\n type: \"text\", \n text: \"Error: This command requires the ValTown CLI (vt) to be installed\"\n }],\n isError: true,\n };\n }\n \n try {\n const createArgs = [\"create\", name, \"--template\", templateRepo, \"--privacy\", privacy];\n \n if (description) {\n createArgs.push(\"--description\", description);\n }\n \n // Add JSON output format\n createArgs.push(\"--json\");\n \n const result = await runVtCommand(createArgs);\n \n if (result.success) {\n try {\n const data = JSON.parse(result.output);\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n };\n } catch {\n return {\n content: [{type: \"text\", text: result.output}],\n };\n }\n } else {\n return {\n content: [{type: \"text\", text: `Error creating project from template: ${result.error}`}],\n isError: true,\n };\n }\n } catch (error) {\n console.error(error);\n return {\n content: [{type: \"text\", text: `Error creating project from template: ${getErrorMessage(error)}`}],\n isError: true,\n };\n }\n }\n );\n \n // Sync all changes in a project\n server.tool(\n \"sync-project\",\n \"Sync all changes in a project (pull and push)\",\n {\n projectId: z.string().describe(\"ID of the project to sync\"),\n branchId: z.string().optional().describe(\"ID of the branch to sync (optional, defaults to main)\"),\n },\n async ({projectId, branchId}: {\n projectId: string\n branchId?: string\n }) => {\n const cliAvailable = await getCliAvailability();\n \n if (!cliAvailable) {\n return {\n content: [{\n type: \"text\", \n text: \"Error: This command requires the ValTown CLI (vt) to be installed\"\n }],\n isError: true,\n };\n }\n \n try {\n // First pull changes\n const pullArgs = [\"pull\", \"--project\", projectId];\n \n if (branchId) {\n pullArgs.push(\"--branch\", branchId);\n }\n \n const pullResult = await runVtCommand(pullArgs);\n \n if (!pullResult.success) {\n return {\n content: [{type: \"text\", text: `Error pulling changes: ${pullResult.error}`}],\n isError: true,\n };\n }\n \n // Then push changes\n const pushArgs = [\"push\", \"--project\", projectId];\n \n if (branchId) {\n pushArgs.push(\"--branch\", branchId);\n }\n \n // Add JSON output format\n pushArgs.push(\"--json\");\n \n const pushResult = await runVtCommand(pushArgs);\n \n if (pushResult.success) {\n try {\n const data = JSON.parse(pushResult.output);\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n };\n } catch {\n return {\n content: [{type: \"text\", text: \"Project synchronized successfully\"}],\n };\n }\n } else {\n return {\n content: [{type: \"text\", text: `Error pushing changes: ${pushResult.error}`}],\n isError: true,\n };\n }\n } catch (error) {\n console.error(error);\n return {\n content: [{type: \"text\", text: `Error syncing project: ${getErrorMessage(error)}`}],\n isError: true,\n };\n }\n }\n );\n \n // Watch a Val for local changes and automatically push them\n server.tool(\n \"watch-val\",\n \"Watch a Val for local changes and automatically push them\",\n {\n valId: z.string().describe(\"ID of the val\"),\n branchId: z.string().optional().describe(\"ID of the branch (optional, defaults to main)\"),\n },\n async ({valId, branchId}) => {\n // Check for CLI availability (required for this tool)\n const cliAvailable = await getCliAvailability();\n \n if (!cliAvailable) {\n return {\n content: [{type: \"text\", text: \"This tool requires the ValTown CLI, which is not available.\"}],\n isError: true,\n };\n }\n\n try {\n // Prepare a workspace with the val cloned\n console.log(`Using CLI to watch val: ${valId}`);\n const workspace = await prepareValWorkspace(valId);\n \n if (!workspace.success || !workspace.workspacePath) {\n return {\n content: [{type: \"text\", text: `Error preparing workspace: ${workspace.error || \"Unknown error\"}`}],\n isError: true,\n };\n }\n \n // If a branch is specified, checkout that branch\n if (branchId) {\n const checkoutResult = await runVtCommand([\"checkout\", branchId]);\n \n if (!checkoutResult.success) {\n await cleanupTempDirectory(workspace.workspacePath);\n return {\n content: [{type: \"text\", text: `Error checking out branch: ${checkoutResult.error}`}],\n isError: true,\n };\n }\n }\n \n // Start watching the val\n // Note: In a real implementation, this would need to be a long-running process\n // This example just initiates the watch command and reports success\n const watchResult = await runVtCommand([\"watch\", \"--json\"]);\n \n // Note: In a real implementation, we would keep the watch process running\n // and handle cleanup when appropriate. For this example, we'll just report success\n \n // Clean up workspace (in a real implementation, this would happen later)\n await cleanupTempDirectory(workspace.workspacePath);\n \n if (!watchResult.success) {\n return {\n content: [{type: \"text\", text: `Error starting watch: ${watchResult.error}`}],\n isError: true,\n };\n }\n \n // Return success message\n return {\n content: [{type: \"text\", text: `Started watching val ${valId}${branchId ? ` (branch: ${branchId})` : ''}.`}],\n };\n } catch (error) {\n console.error(error);\n return {\n content: [{type: \"text\", text: `Error watching val: ${getErrorMessage(error)}`}],\n isError: true,\n };\n }\n }\n );\n \n // Pull latest changes from Val Town\n server.tool(\n \"pull-val\",\n \"Pull latest changes from Val Town\",\n {\n valId: z.string().describe(\"ID of the val\"),\n branchId: z.string().optional().describe(\"ID of the branch (optional, defaults to main)\"),\n },\n async ({valId, branchId}) => {\n // Check for CLI availability (required for this tool)\n const cliAvailable = await getCliAvailability();\n \n if (!cliAvailable) {\n return {\n content: [{type: \"text\", text: \"This tool requires the ValTown CLI, which is not available.\"}],\n isError: true,\n };\n }\n\n try {\n // Prepare a workspace with the val cloned\n console.log(`Using CLI to pull latest changes for val: ${valId}`);\n const workspace = await prepareValWorkspace(valId);\n \n if (!workspace.success || !workspace.workspacePath) {\n return {\n content: [{type: \"text\", text: `Error preparing workspace: ${workspace.error || \"Unknown error\"}`}],\n isError: true,\n };\n }\n \n // If a branch is specified, checkout that branch\n if (branchId) {\n const checkoutResult = await runVtCommand([\"checkout\", branchId]);\n \n if (!checkoutResult.success) {\n await cleanupTempDirectory(workspace.workspacePath);\n return {\n content: [{type: \"text\", text: `Error checking out branch: ${checkoutResult.error}`}],\n isError: true,\n };\n }\n }\n \n // Pull latest changes\n const pullResult = await runVtCommand([\"pull\", \"--json\"]);\n \n // Clean up workspace\n await cleanupTempDirectory(workspace.workspacePath);\n \n if (!pullResult.success) {\n return {\n content: [{type: \"text\", text: `Error pulling changes: ${pullResult.error}`}],\n isError: true,\n };\n }\n \n // Parse and return result\n const parsedOutput = parseCliJsonOutput(pullResult.output);\n if (parsedOutput) {\n return {\n content: [{type: \"text\", text: JSON.stringify(parsedOutput, null, 2)}],\n };\n }\n \n // Fallback success message\n return {\n content: [{type: \"text\", text: `Successfully pulled latest changes for val ${valId}${branchId ? ` (branch: ${branchId})` : ''}.`}],\n };\n } catch (error) {\n console.error(error);\n return {\n content: [{type: \"text\", text: `Error pulling changes: ${getErrorMessage(error)}`}],\n isError: true,\n };\n }\n }\n );\n \n // Open a Val in the default web browser\n server.tool(\n \"browse-val\",\n \"Open a Val in the default web browser\",\n {\n valId: z.string().describe(\"ID of the val\"),\n },\n async ({valId}) => {\n // Check for CLI availability (required for this tool)\n const cliAvailable = await getCliAvailability();\n \n if (!cliAvailable) {\n return {\n content: [{type: \"text\", text: \"This tool requires the ValTown CLI, which is not available.\"}],\n isError: true,\n };\n }\n\n try {\n // Note: This is typically not an operation you'd want to do from a server\n // as it would open a browser on the server machine. In a real implementation,\n // you might just return the URL to open.\n console.log(`Using CLI to get browse URL for val: ${valId}`);\n \n // Run the command to get the URL\n const browseResult = await runVtCommand([\"browse\", valId, \"--dry-run\", \"--json\"]);\n \n if (!browseResult.success) {\n return {\n content: [{type: \"text\", text: `Error getting browse URL: ${browseResult.error}`}],\n isError: true,\n };\n }\n \n // Parse and return URL\n const parsedOutput = parseCliJsonOutput(browseResult.output);\n if (parsedOutput) {\n return {\n content: [{type: \"text\", text: JSON.stringify(parsedOutput, null, 2)}],\n };\n }\n \n // Fallback URL format\n return {\n content: [{type: \"text\", text: `https://www.val.town/v/${valId}`}],\n };\n } catch (error) {\n console.error(error);\n return {\n content: [{type: \"text\", text: `Error getting browse URL: ${getErrorMessage(error)}`}],\n isError: true,\n };\n }\n }\n );\n \n // Remix an existing Val to create a new one\n server.tool(\n \"remix-val\",\n \"Remix an existing Val to create a new one\",\n {\n sourceValUri: z.string().describe(\"URI of the source Val to remix (username/valname)\"),\n newValName: z.string().describe(\"Name for the new Val\"),\n privacy: z.enum([\"public\", \"unlisted\", \"private\"]).default(\"public\").describe(\"Privacy setting for the new Val\"),\n },\n async ({sourceValUri, newValName, privacy}) => {\n // Check for CLI availability (required for this tool)\n const cliAvailable = await getCliAvailability();\n \n if (!cliAvailable) {\n return {\n content: [{type: \"text\", text: \"This tool requires the ValTown CLI, which is not available.\"}],\n isError: true,\n };\n }\n\n try {\n console.log(`Using CLI to remix val: ${sourceValUri} to ${newValName}`);\n \n // Build command arguments\n const args = [\"remix\", sourceValUri, newValName, \"--json\"];\n \n // Add privacy option if not public (public is default)\n if (privacy !== \"public\") {\n args.push(\"--privacy\", privacy);\n }\n \n // Run the remix command\n const remixResult = await runVtCommand(args);\n \n if (!remixResult.success) {\n return {\n content: [{type: \"text\", text: `Error remixing val: ${remixResult.error}`}],\n isError: true,\n };\n }\n \n // Parse and return result\n const parsedOutput = parseCliJsonOutput(remixResult.output);\n if (parsedOutput) {\n return {\n content: [{type: \"text\", text: JSON.stringify(parsedOutput, null, 2)}],\n };\n }\n \n // Fallback success message\n return {\n content: [{type: \"text\", text: `Successfully remixed ${sourceValUri} to ${newValName}.`}],\n };\n } catch (error) {\n console.error(error);\n return {\n content: [{type: \"text\", text: `Error remixing val: ${getErrorMessage(error)}`}],\n isError: true,\n };\n }\n }\n );\n \n // Show the working tree status of a Val\n server.tool(\n \"get-val-status\",\n \"Show the working tree status of a Val\",\n {\n valId: z.string().describe(\"ID of the val\"),\n branchId: z.string().optional().describe(\"ID of the branch (optional, defaults to main)\"),\n },\n async ({valId, branchId}) => {\n // Check for CLI availability (required for this tool)\n const cliAvailable = await getCliAvailability();\n \n if (!cliAvailable) {\n return {\n content: [{type: \"text\", text: \"This tool requires the ValTown CLI, which is not available.\"}],\n isError: true,\n };\n }\n\n try {\n // Prepare a workspace with the val cloned\n console.log(`Using CLI to get status for val: ${valId}`);\n const workspace = await prepareValWorkspace(valId);\n \n if (!workspace.success || !workspace.workspacePath) {\n return {\n content: [{type: \"text\", text: `Error preparing workspace: ${workspace.error || \"Unknown error\"}`}],\n isError: true,\n };\n }\n \n // If a branch is specified, checkout that branch\n if (branchId) {\n const checkoutResult = await runVtCommand([\"checkout\", branchId]);\n \n if (!checkoutResult.success) {\n await cleanupTempDirectory(workspace.workspacePath);\n return {\n content: [{type: \"text\", text: `Error checking out branch: ${checkoutResult.error}`}],\n isError: true,\n };\n }\n }\n \n // Get status\n const statusResult = await runVtCommand([\"status\", \"--json\"]);\n \n // Clean up workspace\n await cleanupTempDirectory(workspace.workspacePath);\n \n if (!statusResult.success) {\n return {\n content: [{type: \"text\", text: `Error getting status: ${statusResult.error}`}],\n isError: true,\n };\n }\n \n // Parse and return result\n const parsedOutput = parseCliJsonOutput(statusResult.output);\n if (parsedOutput) {\n return {\n content: [{type: \"text\", text: JSON.stringify(parsedOutput, null, 2)}],\n };\n }\n \n // Fallback status message\n return {\n content: [{type: \"text\", text: statusResult.output}],\n };\n } catch (error) {\n console.error(error);\n return {\n content: [{type: \"text\", text: `Error getting val status: ${getErrorMessage(error)}`}],\n isError: true,\n };\n }\n }\n );\n}",[],[278,300],{"_105":307,"_95":308,"_157":309,"_159":279,"_160":140,"_161":120,"_162":163,"_107":310,"_101":32,"_165":190,"_167":311,"_168":22,"_169":-5,"_170":312,"_99":313},"d7337b14-77b0-11f0-a503-0224a6c84d84","fileTools.ts","01989fb6-5dc4-7132-98d1-0899cab2939d",["D",1755026185672],"import {McpServer} from \"npm:@modelcontextprotocol/sdk/server/mcp.js\"\nimport {Config} from \"../lib/types.ts\"\nimport {callValTownApi} from \"../lib/api.ts\"\nimport {getErrorMessage} from \"../lib/errorUtils.ts\"\nimport {getCliAvailability, runVtCommand, parseCliJsonOutput, prepareValWorkspace, cleanupTempDirectory} from \"../lib/vtCli.ts\"\nimport * as path from \"jsr:@std/path\"\nimport {z} from \"npm:zod\"\n\n/**\n * Helper to perform file operations using the CLI\n * This handles the common pattern of:\n * 1. Clone the val to a workspace\n * 2. Checkout the correct branch if needed\n * 3. Modify files on disk\n * 4. Push changes back to Val Town\n * 5. Clean up workspace\n */\nasync function performFileOperationWithCli(\n config: Config,\n valId: string,\n branchId: string | undefined,\n operation: (workspacePath: string) => Promise,\n operationDescription: string\n): Promise<{success: boolean; output?: any; error?: string}> {\n try {\n console.log(`Using CLI for ${operationDescription} on val: ${valId}`)\n\n // Prepare workspace with the val cloned\n const workspace = await prepareValWorkspace(valId)\n\n if (!workspace.success || !workspace.workspacePath) {\n return {\n success: false,\n error: workspace.error || \"Failed to prepare workspace\"\n }\n }\n\n // If a branch is specified, checkout that branch\n if (branchId) {\n const checkoutResult = await runVtCommand([\"checkout\", branchId])\n\n if (!checkoutResult.success) {\n await cleanupTempDirectory(workspace.workspacePath!)\n return {\n success: false,\n error: `Failed to checkout branch: ${checkoutResult.error}`\n }\n }\n }\n\n // Perform the specific file operation\n const operationSuccess = await operation(workspace.workspacePath!)\n\n if (!operationSuccess) {\n await cleanupTempDirectory(workspace.workspacePath!)\n return {\n success: false,\n error: `File operation failed`\n }\n }\n\n // Push changes back to Val Town\n const pushResult = await runVtCommand([\"push\", \"--json\"])\n\n // Clean up temporary workspace\n await cleanupTempDirectory(workspace.workspacePath!)\n\n if (!pushResult.success) {\n return {\n success: false,\n error: `Failed to push changes: ${pushResult.error}`\n }\n }\n\n // Parse and return the result\n const parsedOutput = parseCliJsonOutput(pushResult.output)\n return {\n success: true,\n output: parsedOutput\n }\n } catch (error) {\n console.error(`CLI error during ${operationDescription}:`, getErrorMessage(error))\n return {\n success: false,\n error: getErrorMessage(error)\n }\n }\n}\n\nexport function registerFileTools(server: McpServer, config: Config) {\n // List files in a val branch\n server.tool(\n \"list-files\",\n \"List files in a val branch\",\n {\n valId: z.string().describe(\"ID of the val\"),\n path: z.string().default(\"\").describe(\"Path to directory (leave empty for root)\"),\n branchId: z.string().optional().describe(\"ID of the branch (optional, defaults to main)\"),\n recursive: z.boolean().default(false).describe(\"Whether to list files recursively\"),\n limit: z.number().int().min(1).max(100).default(20).describe(\"Maximum number of results to return\"),\n offset: z.number().int().min(0).default(0).describe(\"Number of items to skip for pagination\"),\n },\n async ({valId, path, branchId, recursive, limit, offset}) => {\n // Check for CLI preference\n const useCliIfAvailable = config.cli?.preferCli ?? false\n const cliAvailable = useCliIfAvailable && await getCliAvailability()\n\n if (cliAvailable) {\n try {\n // The CLI status command can be used to list files\n console.log(`Using CLI to list files for val: ${valId}`)\n\n // Prepare a workspace with the val cloned\n const workspace = await prepareValWorkspace(valId)\n\n if (workspace.success && workspace.workspacePath) {\n // If a branch is specified, checkout that branch first\n if (branchId) {\n const checkoutResult = await runVtCommand([\"checkout\", branchId])\n\n if (!checkoutResult.success) {\n console.error(`Failed to checkout branch: ${checkoutResult.error}`)\n await cleanupTempDirectory(workspace.workspacePath!)\n // Fall back to API\n console.error(\"CLI error when checking out branch, falling back to API\")\n throw new Error(\"Failed to checkout branch\")\n }\n }\n\n // Run the status command to list files\n const result = await runVtCommand([\"status\", \"--json\"])\n\n // Clean up the temporary directory\n await cleanupTempDirectory(workspace.workspacePath!)\n\n if (result.success) {\n // Parse JSON output\n const parsedOutput = parseCliJsonOutput(result.output)\n if (parsedOutput) {\n // Filter by path if needed\n // Apply pagination manually if needed\n return {\n content: [{type: \"text\", text: JSON.stringify(parsedOutput, null, 2)}],\n }\n }\n }\n }\n\n console.error(`CLI error when listing files, falling back to API: ${workspace.error || \"Unknown error\"}`)\n // Fall back to API on error\n } catch (error) {\n console.error(\"CLI error, falling back to API:\", getErrorMessage(error))\n // Fall back to API on error\n }\n }\n\n // API implementation (original code)\n try {\n let queryParams = `?path=${encodeURIComponent(path)}&recursive=${recursive}&limit=${limit}&offset=${offset}`\n if (branchId) {\n queryParams += `&branch_id=${encodeURIComponent(branchId)}`\n }\n\n const data = await callValTownApi(\n config,\n `/v2/vals/${valId}/files${queryParams}`\n )\n\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error listing files: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n\n // Get file content from a val branch\n server.tool(\n \"get-file\",\n \"Get file content from a val branch\",\n {\n valId: z.string().describe(\"ID of the val\"),\n path: z.string().describe(\"Path to the file\"),\n branchId: z.string().optional().describe(\"ID of the branch (optional, defaults to main)\"),\n },\n async ({valId, path, branchId}) => {\n try {\n let queryParams = `?path=${encodeURIComponent(path)}`\n if (branchId) {\n queryParams += `&branch_id=${encodeURIComponent(branchId)}`\n }\n\n const response = await fetch(\n `${config.apiBase}/v2/vals/${valId}/files/content${queryParams}`,\n {\n headers: {\n 'Authorization': `Bearer ${config.apiToken}`,\n },\n }\n )\n\n if (!response.ok) {\n const errorText = await response.text()\n throw new Error(`API error (${response.status}): ${errorText}`)\n }\n\n const content = await response.text()\n\n return {\n content: [{type: \"text\", text: content}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error getting file content: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n\n // Create a new file or directory in a val branch\n server.tool(\n \"create-file-or-directory\",\n \"Create a new file or directory in a val branch\",\n {\n valId: z.string().describe(\"ID of the val\"),\n path: z.string().describe(\"Path to the new file or directory\"),\n type: z.enum([\"file\", \"interval\", \"http\", \"email\", \"script\", \"directory\"]).describe(\"Type of resource to create: file, interval, http, email, script, or directory. Only use `file` for non code files. For code files, including scripts and UI Code, use `interval`, `http`, `email`, or `script`.\"),\n content: z.string().optional().describe(\"Content for the file (required for files, not for directories, the code must be in typescript)\"),\n branchId: z.string().optional().describe(\"ID of the branch (optional, defaults to main)\"),\n },\n async ({valId, path: filePath, type, content, branchId}) => {\n // Check for CLI preference\n const useCliIfAvailable = config.cli?.preferCli ?? false\n const cliAvailable = useCliIfAvailable && await getCliAvailability()\n\n if (cliAvailable) {\n try {\n const result = await performFileOperationWithCli(\n config,\n valId,\n branchId,\n async (workspacePath) => {\n try {\n const fullPath = path.join(workspacePath, filePath)\n\n // Create directory if needed\n if (type === \"directory\") {\n try {\n await Deno.mkdir(fullPath, {recursive: true})\n return true\n } catch (error) {\n console.error(`Failed to create directory: ${getErrorMessage(error)}`)\n return false\n }\n } else {\n // Create parent directories if needed\n const parentDir = path.dirname(fullPath)\n try {\n await Deno.mkdir(parentDir, {recursive: true})\n } catch (error) {\n // Ignore if directory already exists\n if (!(error instanceof Deno.errors.AlreadyExists)) {\n console.error(`Failed to create parent directory: ${getErrorMessage(error)}`)\n return false\n }\n }\n\n // Create file with content\n try {\n await Deno.writeTextFile(fullPath, content || \"\")\n return true\n } catch (error) {\n console.error(`Failed to create file: ${getErrorMessage(error)}`)\n return false\n }\n }\n } catch (error) {\n console.error(`File operation error: ${getErrorMessage(error)}`)\n return false\n }\n },\n `creating ${type === \"directory\" ? \"directory\" : \"file\"} at ${filePath}`\n )\n\n if (result.success) {\n return {\n content: [{type: \"text\", text: JSON.stringify(result.output || `Created ${type} at ${filePath}`, null, 2)}],\n }\n } else {\n console.error(`CLI error when creating ${type}, falling back to API: ${result.error}`)\n // Fall back to API\n }\n } catch (error) {\n console.error(\"CLI error, falling back to API:\", getErrorMessage(error))\n // Fall back to API on error\n }\n }\n\n // API implementation (original code)\n try {\n let queryParams = `?path=${encodeURIComponent(filePath)}`\n if (branchId) {\n queryParams += `&branch_id=${encodeURIComponent(branchId)}`\n }\n\n const requestBody = {\n type,\n ...(type !== \"directory\" ? {content: content || \"\"} : {content: null}),\n }\n\n const data = await callValTownApi(\n config,\n `/v2/vals/${valId}/files${queryParams}`,\n {\n method: \"POST\",\n body: JSON.stringify(requestBody),\n }\n )\n\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error creating file or directory: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n\n // Update an existing file in a val branch\n server.tool(\n \"update-file\",\n \"Update an existing file in a val branch\",\n {\n valId: z.string().describe(\"ID of the val\"),\n path: z.string().describe(\"Path to the file\"),\n content: z.string().describe(\"New content for the file\"),\n branchId: z.string().optional().describe(\"ID of the branch (optional, defaults to main)\"),\n },\n async ({valId, path: filePath, content, branchId}) => {\n // Check for CLI preference\n const useCliIfAvailable = config.cli?.preferCli ?? false\n const cliAvailable = useCliIfAvailable && await getCliAvailability()\n\n if (cliAvailable) {\n try {\n const result = await performFileOperationWithCli(\n config,\n valId,\n branchId,\n async (workspacePath) => {\n try {\n const fullPath = path.join(workspacePath, filePath)\n\n // Update file with new content\n try {\n await Deno.writeTextFile(fullPath, content)\n return true\n } catch (error) {\n console.error(`Failed to update file: ${getErrorMessage(error)}`)\n return false\n }\n } catch (error) {\n console.error(`File operation error: ${getErrorMessage(error)}`)\n return false\n }\n },\n `updating file at ${filePath}`\n )\n\n if (result.success) {\n return {\n content: [{type: \"text\", text: JSON.stringify(result.output || `Updated file at ${filePath}`, null, 2)}],\n }\n } else {\n console.error(`CLI error when updating file, falling back to API: ${result.error}`)\n // Fall back to API\n }\n } catch (error) {\n console.error(\"CLI error, falling back to API:\", getErrorMessage(error))\n // Fall back to API on error\n }\n }\n\n // API implementation (original code)\n try {\n let queryParams = `?path=${encodeURIComponent(filePath)}`\n if (branchId) {\n queryParams += `&branch_id=${encodeURIComponent(branchId)}`\n }\n\n const data = await callValTownApi(\n config,\n `/v2/vals/${valId}/files${queryParams}`,\n {\n method: \"PUT\",\n body: JSON.stringify({content}),\n }\n )\n\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error updating file: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n\n // Delete a file or directory from a val branch\n server.tool(\n \"delete-file-or-directory\",\n \"Delete a file or directory from a val branch\",\n {\n valId: z.string().describe(\"ID of the val\"),\n path: z.string().describe(\"Path to the file or directory\"),\n recursive: z.boolean().default(false).describe(\"Whether to recursively delete all files in the directory\"),\n branchId: z.string().optional().describe(\"ID of the branch (optional, defaults to main)\"),\n },\n async ({valId, path: filePath, recursive, branchId}) => {\n // Check for CLI preference\n const useCliIfAvailable = config.cli?.preferCli ?? false\n const cliAvailable = useCliIfAvailable && await getCliAvailability()\n\n if (cliAvailable) {\n try {\n const result = await performFileOperationWithCli(\n config,\n valId,\n branchId,\n async (workspacePath) => {\n try {\n const fullPath = path.join(workspacePath, filePath)\n\n // Check if it's a file or directory\n try {\n const stat = await Deno.stat(fullPath)\n\n if (stat.isDirectory) {\n // Delete directory\n try {\n await Deno.remove(fullPath, {recursive})\n return true\n } catch (error) {\n\n console.error(`Failed to delete directory: ${getErrorMessage(error)}`)\n return false\n }\n } else {\n // Delete file\n try {\n await Deno.remove(fullPath)\n return true\n } catch (error) {\n console.error(`Failed to delete file: ${getErrorMessage(error)}`)\n return false\n }\n }\n } catch (error) {\n console.error(`Failed to check path: ${getErrorMessage(error)}`)\n return false\n }\n } catch (error) {\n console.error(`File operation error: ${getErrorMessage(error)}`)\n return false\n }\n },\n `deleting path ${filePath}`\n )\n\n if (result.success) {\n return {\n content: [{type: \"text\", text: `File or directory at path ${filePath} deleted successfully.`}],\n }\n } else {\n console.error(`CLI error when deleting path, falling back to API: ${result.error}`)\n // Fall back to API\n }\n } catch (error) {\n console.error(\"CLI error, falling back to API:\", getErrorMessage(error))\n // Fall back to API on error\n }\n }\n\n // API implementation (original code)\n try {\n let queryParams = `?path=${encodeURIComponent(filePath)}&recursive=${recursive}`\n if (branchId) {\n queryParams += `&branch_id=${encodeURIComponent(branchId)}`\n }\n\n await callValTownApi(\n config,\n `/v2/vals/${valId}/files${queryParams}`,\n {\n method: \"DELETE\",\n }\n )\n\n return {\n content: [{type: \"text\", text: `File or directory at path ${filePath} deleted successfully.`}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error deleting file or directory: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n}",[],[278,308],{"_105":315,"_95":316,"_157":317,"_159":279,"_160":140,"_161":120,"_162":163,"_107":318,"_101":32,"_165":190,"_167":319,"_168":22,"_169":-5,"_170":320,"_99":321},"d7338758-77b0-11f0-a503-0224a6c84d84","projectTools.ts","01989fb6-5dc4-7132-98d1-0d516ab2f8a2",["D",1755026185672],"import {McpServer} from \"npm:@modelcontextprotocol/sdk/server/mcp.js\"\nimport {z} from \"npm:zod\"\nimport {Config} from \"../lib/types.ts\"\nimport {callValTownApi} from \"../lib/api.ts\"\nimport {getCliAvailability, runVtCommand} from \"../lib/vtCli.ts\"\n\nexport function registerProjectTools(server: McpServer, config: Config) {\n // List all projects\n server.tool(\n \"list-projects\",\n \"List all projects accessible to the current user\",\n {\n limit: z.number().int().min(1).max(100).default(20).describe(\"Maximum number of results to return\"),\n offset: z.number().int().min(0).default(0).describe(\"Number of items to skip for pagination\"),\n },\n async ({limit, offset}: {limit: number; offset: number}) => {\n try {\n const data = await callValTownApi(\n config,\n `/v1/me/projects?limit=${limit}&offset=${offset}`\n )\n\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n const errorMessage = error instanceof Error ? error.message : String(error)\n return {\n content: [{type: \"text\", text: `Error listing projects: ${errorMessage}`}],\n isError: true,\n }\n }\n }\n )\n\n // Get a specific project\n server.tool(\n \"get-project\",\n \"Get details about a specific project\",\n {\n projectId: z.string().describe(\"ID of the project\"),\n },\n async ({projectId}: {projectId: string}) => {\n try {\n const data = await callValTownApi(config, `/v1/projects/${projectId}`)\n\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n const errorMessage = error instanceof Error ? error.message : String(error)\n return {\n content: [{type: \"text\", text: `Error getting project: ${errorMessage}`}],\n isError: true,\n }\n }\n }\n )\n\n // Get a project by username and project name\n server.tool(\n \"get-project-by-name\",\n \"Get a project by username and project name\",\n {\n username: z.string().describe(\"Username of the project's owner\"),\n projectName: z.string().describe(\"Name of the project\"),\n },\n async ({username, projectName}: {username: string; projectName: string}) => {\n // Check if CLI is available\n const cliAvailable = await getCliAvailability()\n\n if (cliAvailable && (config.cli?.preferCli ?? false)) {\n try {\n // First, try to use vt CLI\n const result = await runVtCommand([\"clone\", `${username}/${projectName}`, \"--json\"])\n\n if (result.success) {\n // Parse JSON output from CLI\n const data = JSON.parse(result.output)\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n }\n // If CLI fails, fall back to API\n } catch (error) {\n console.error(\"CLI error:\", error)\n // Continue to API fallback\n }\n }\n\n // Fallback to original API implementation\n try {\n const data = await callValTownApi(\n config,\n `/v1/alias/projects/${encodeURIComponent(username)}/${encodeURIComponent(projectName)}`\n )\n\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n const errorMessage = error instanceof Error ? error.message : String(error)\n return {\n content: [{type: \"text\", text: `Error getting project: ${errorMessage}`}],\n isError: true,\n }\n }\n }\n )\n\n // Create a new project\n server.tool(\n \"create-project\",\n \"Create a new project\",\n {\n name: z.string()\n .regex(/^[a-zA-Z][a-zA-Z0-9_]*$/, \"Project name must start with a letter and can only contain letters, numbers, and underscores\")\n .describe(\"Name for the project\"),\n privacy: z.enum([\"public\", \"unlisted\", \"private\"]).default(\"unlisted\")\n .describe(\"Privacy setting: public, unlisted, or private\"),\n description: z.string().max(64).optional()\n .describe(\"Description for the project (optional, max 64 characters)\"),\n imageUrl: z.string().optional().describe(\"URL to an image for the project (optional)\"),\n },\n async ({name, privacy, description, imageUrl}: {\n name: string\n privacy: \"public\" | \"unlisted\" | \"private\"\n description?: string\n imageUrl?: string\n }) => {\n try {\n // Validate project name\n if (!/^[a-zA-Z][a-zA-Z0-9_]*$/.test(name)) {\n return {\n content: [{\n type: \"text\",\n text: \"Error: Project name must start with a letter and can only contain letters, numbers, and underscores\"\n }],\n isError: true,\n }\n }\n\n // Validate description length\n if (description && description.length > 64) {\n return {\n content: [{\n type: \"text\",\n text: \"Error: Description must be 64 characters or less\"\n }],\n isError: true,\n }\n }\n\n const payload = {\n name,\n privacy,\n ...(description ? {description} : {}),\n ...(imageUrl ? {imageUrl} : {}),\n }\n\n const data = await callValTownApi(config, \"/v1/projects\", {\n method: \"POST\",\n body: JSON.stringify(payload),\n })\n\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n const errorMessage = error instanceof Error ? error.message : String(error)\n return {\n content: [{type: \"text\", text: `Error creating project: ${errorMessage}`}],\n isError: true,\n }\n }\n }\n )\n\n // Delete a project\n server.tool(\n \"delete-project\",\n \"Delete a project by ID\",\n {\n projectId: z.string().describe(\"ID of the project to delete\"),\n },\n async ({projectId}: {projectId: string}) => {\n try {\n await callValTownApi(config, `/v1/projects/${projectId}`, {\n method: \"DELETE\",\n })\n\n return {\n content: [{type: \"text\", text: \"Project deleted successfully\"}],\n }\n } catch (error) {\n console.error(error)\n const errorMessage = error instanceof Error ? error.message : String(error)\n return {\n content: [{type: \"text\", text: `Error deleting project: ${errorMessage}`}],\n isError: true,\n }\n }\n }\n )\n}",[],[278,316],{"_105":323,"_95":324,"_157":325,"_159":279,"_160":140,"_161":120,"_162":163,"_107":326,"_101":32,"_165":190,"_167":327,"_168":22,"_169":-5,"_170":328,"_99":329},"d7338c26-77b0-11f0-a503-0224a6c84d84","sqliteTools.ts","01989fb6-5dc4-7132-98d1-10a5a235441d",["D",1755026185672],"import {McpServer} from \"npm:@modelcontextprotocol/sdk/server/mcp.js\"\nimport {z} from \"npm:zod\"\nimport {Config} from \"../lib/types.ts\"\nimport {callValTownApi} from \"../lib/api.ts\"\nimport {getErrorMessage} from \"../lib/errorUtils.ts\"\n\nexport function registerSqliteTools(server: McpServer, config: Config) {\n // Execute SQL\n server.tool(\n \"execute-sql\",\n \"Execute a SQL query against a SQLite database\",\n {\n statement: z.string().describe(\"SQL statement to execute\"),\n },\n async ({statement}: {statement: string}) => {\n try {\n const data = await callValTownApi(config, \"/v1/sqlite/execute\", {\n method: \"POST\",\n body: JSON.stringify({statement}),\n })\n\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error executing SQL: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n\n // Execute SQL batch\n server.tool(\n \"execute-sql-batch\",\n \"Execute multiple SQL statements against a SQLite database\",\n {\n statements: z.array(z.string()).describe(\"Array of SQL statements to execute\"),\n mode: z.enum([\"read\", \"write\"]).default(\"read\").describe(\"Mode of the statements (read or write)\"),\n },\n async ({statements, mode}: {\n statements: string[]\n mode: \"read\" | \"write\"\n }) => {\n try {\n const data = await callValTownApi(config, \"/v1/sqlite/batch\", {\n method: \"POST\",\n body: JSON.stringify({statements, mode}),\n })\n\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error executing SQL batch: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n\n // Fix for sqlite-query function\n server.tool(\n \"sqlite-query\",\n \"Execute a SQL query against a SQLite database\",\n {\n statement: z.string().describe(\"SQL statement to execute\"),\n },\n async ({statement}: {statement: string}) => {\n try {\n const data = await callValTownApi(config, \"/v1/sqlite/query\", {\n method: \"POST\",\n body: JSON.stringify({statement}),\n })\n\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error executing SQL query: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n\n // Fix for sqlite-exec function\n server.tool(\n \"sqlite-exec\",\n \"Execute multiple SQL statements against a SQLite database\",\n {\n statements: z.array(z.string()).describe(\"Array of SQL statements to execute\"),\n mode: z.enum([\"read\", \"write\"]).default(\"read\").describe(\"Mode of the statements (read or write)\"),\n },\n async ({statements, mode}: {\n statements: string[]\n mode: \"read\" | \"write\" // or a more specific type if applicable\n }) => {\n try {\n const data = await callValTownApi(config, \"/v1/sqlite/exec\", {\n method: \"POST\",\n body: JSON.stringify({statements, mode}),\n })\n\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error executing SQL exec: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n}",[],[278,324],{"_105":331,"_95":332,"_157":333,"_159":279,"_160":140,"_161":120,"_162":163,"_107":334,"_101":32,"_165":190,"_167":335,"_168":22,"_169":-5,"_170":336,"_99":337},"d7338e74-77b0-11f0-a503-0224a6c84d84","userTools.ts","01989fb6-5dc4-7132-98d1-14953fd18a13",["D",1755026185672],"import {McpServer} from \"npm:@modelcontextprotocol/sdk/server/mcp.js\"\nimport {z} from \"npm:zod\"\nimport {Config} from \"../lib/types.ts\"\nimport {callValTownApi} from \"../lib/api.ts\"\nimport {getErrorMessage} from \"../lib/errorUtils.ts\"\n\nexport function registerUserTools(server: McpServer, config: Config) {\n // Get user information by username\n server.tool(\n \"get-user\",\n \"Get basic details about a user by username\",\n {\n username: z.string().describe(\"Username of the user to look for (without @ symbol)\"),\n },\n async ({username}: {username: string}) => {\n try {\n const data = await callValTownApi(config, `/v1/alias/${encodeURIComponent(username)}`)\n\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error getting user: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n\n // Get current user information (me)\n server.tool(\n \"get-my-info\",\n \"Get profile information for the current authenticated user\",\n {},\n async () => {\n try {\n const data = await callValTownApi(config, \"/v1/me\")\n\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error getting user info: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n}",[],[278,332],{"_105":339,"_95":340,"_157":341,"_159":279,"_160":140,"_161":120,"_162":163,"_107":342,"_101":32,"_165":190,"_167":343,"_168":22,"_169":-5,"_170":344,"_99":345},"d73391da-77b0-11f0-a503-0224a6c84d84","valsTools.ts","01989fb6-5dc4-7132-98d1-1bd3bbcd0389",["D",1755026185672],"import {McpServer} from \"npm:@modelcontextprotocol/sdk/server/mcp.js\"\nimport {Config} from \"../lib/types.ts\"\nimport {callValTownApi} from \"../lib/api.ts\"\nimport {getErrorMessage} from \"../lib/errorUtils.ts\"\nimport {getCliAvailability, runVtCommand, parseCliJsonOutput} from \"../lib/vtCli.ts\"\nimport {z} from \"npm:zod\"\n\nexport function registerValTools(server: McpServer, config: Config) {\n // Get val by username and val name\n server.tool(\n \"get-val\",\n \"Get a val by username and val name\",\n {\n username: z.string().describe(\"Username of the val's owner\"),\n valName: z.string().describe(\"Name of the val\"),\n },\n async ({username, valName}) => {\n // Check for CLI preference\n const useCliIfAvailable = config.cli?.preferCli ?? false;\n const cliAvailable = useCliIfAvailable && await getCliAvailability();\n const valReference = `${username}/${valName}`;\n\n if (cliAvailable) {\n try {\n // Use CLI implementation with the clone command and --json flag\n console.log(`Using CLI to get val: ${valReference}`);\n const result = await runVtCommand([\"clone\", valReference, \"--json\"]);\n \n if (result.success) {\n // Parse JSON output\n const parsedOutput = parseCliJsonOutput(result.output);\n if (parsedOutput) {\n return {\n content: [{type: \"text\", text: JSON.stringify(parsedOutput, null, 2)}],\n };\n }\n }\n \n console.error(`CLI error when getting val, falling back to API: ${result.error}`);\n // Fall back to API on error\n } catch (error) {\n console.error(\"CLI error, falling back to API:\", getErrorMessage(error));\n // Fall back to API on error\n }\n }\n\n // API implementation (original code)\n try {\n const data = await callValTownApi(\n config,\n `/v2/alias/vals/${encodeURIComponent(username)}/${encodeURIComponent(valName)}`\n );\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n };\n } catch (error) {\n console.error(error);\n return {\n content: [{type: \"text\", text: `Error getting val: ${getErrorMessage(error)}`}],\n isError: true,\n };\n }\n }\n )\n\n // Search vals\n server.tool(\n \"search-vals\",\n \"Search for vals across the Val Town platform\",\n {\n query: z.string().describe(\"Search query\").min(1).max(256),\n limit: z.number().int().min(1).max(100).default(20).describe(\"Maximum number of results to return\"),\n offset: z.number().int().min(0).default(0).describe(\"Number of items to skip for pagination\"),\n },\n async ({query, limit, offset}) => {\n try {\n const data = await callValTownApi(\n config,\n `/v1/search/vals?query=${encodeURIComponent(query)}&limit=${limit}&offset=${offset}`\n )\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error searching vals: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n\n // Create a new val\n server.tool(\n \"create-val\",\n \"Create a new val (code snippet)\",\n {\n name: z.string().describe(\"Name for the val\"),\n description: z.string().optional().describe(\"Description for the val (optional)\"),\n privacy: z.enum([\"public\", \"unlisted\", \"private\"]).default(\"public\").describe(\"Privacy setting: public, unlisted, or private\"),\n },\n async ({name, description, privacy}) => {\n // Check for CLI preference\n const useCliIfAvailable = config.cli?.preferCli ?? false;\n const cliAvailable = useCliIfAvailable && await getCliAvailability();\n\n if (cliAvailable) {\n try {\n // Use CLI implementation with the create command and --json flag\n console.log(`Using CLI to create val: ${name}`);\n // Build CLI arguments\n const cliArgs = [\"create\", name, \"--json\"];\n \n // Add privacy option if not public (public is default)\n if (privacy !== \"public\") {\n cliArgs.push(\"--privacy\", privacy);\n }\n \n // Add description if provided\n if (description) {\n cliArgs.push(\"--description\", description);\n }\n \n const result = await runVtCommand(cliArgs);\n \n if (result.success) {\n // Parse JSON output\n const parsedOutput = parseCliJsonOutput(result.output);\n if (parsedOutput) {\n return {\n content: [{type: \"text\", text: JSON.stringify(parsedOutput, null, 2)}],\n };\n }\n }\n \n console.error(`CLI error when creating val, falling back to API: ${result.error}`);\n // Fall back to API on error\n } catch (error) {\n console.error(\"CLI error, falling back to API:\", getErrorMessage(error));\n // Fall back to API on error\n }\n }\n\n // API implementation (original code)\n try {\n const requestBody = {\n name,\n privacy,\n ...(description ? {description} : {})\n }\n\n const data = await callValTownApi(\n config,\n `/v2/vals`,\n {\n method: \"POST\",\n body: JSON.stringify(requestBody),\n }\n )\n\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error creating val: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n\n // Delete a val\n server.tool(\n \"delete-val\",\n \"Delete a val by ID\",\n {\n valId: z.string().uuid().describe(\"ID of the val to delete\"),\n },\n async ({valId}) => {\n // Check for CLI preference\n const useCliIfAvailable = config.cli?.preferCli ?? false;\n const cliAvailable = useCliIfAvailable && await getCliAvailability();\n\n if (cliAvailable) {\n try {\n // The CLI delete command requires working in a Val directory, \n // so we need to clone first, then delete\n console.log(`Using CLI to delete val: ${valId}`);\n \n // Use prepareValWorkspace first (would need to implement special workspace setup)\n // For now, we'll use the API implementation instead of complex workspace management\n \n // This could be implemented with temporary directory setup if needed, \n // but for now we'll use the API for deletion as it's simpler\n console.log(\"Deletion via CLI requires workspace setup, using API instead\");\n } catch (error) {\n console.error(\"CLI error, falling back to API:\", getErrorMessage(error));\n // Fall back to API on error\n }\n }\n\n // API implementation (original code)\n try {\n await callValTownApi(\n config,\n `/v2/vals/${valId}`,\n {\n method: \"DELETE\",\n }\n )\n\n return {\n content: [{type: \"text\", text: `Val ${valId} deleted successfully.`}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error deleting val: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n\n // List all vals for authenticated user\n server.tool(\n \"list-my-vals\",\n \"List all vals for the authenticated user\",\n {\n limit: z.number().int().min(1).max(100).default(20).describe(\"Maximum number of results to return\"),\n offset: z.number().int().min(0).default(0).describe(\"Number of items to skip for pagination\"),\n },\n async ({limit, offset}) => {\n // Check for CLI preference\n const useCliIfAvailable = config.cli?.preferCli ?? false;\n const cliAvailable = useCliIfAvailable && await getCliAvailability();\n\n if (cliAvailable) {\n try {\n // Use CLI implementation with the list command and --json flag\n console.log(`Using CLI to list vals`);\n \n const cliArgs = [\"list\", \"--json\"];\n // Note: the CLI may not support limit and offset params\n // We could add support for --limit and --offset if the CLI adds these options\n \n const result = await runVtCommand(cliArgs);\n \n if (result.success) {\n // Parse JSON output\n const parsedOutput = parseCliJsonOutput(result.output);\n if (parsedOutput) {\n // Apply limit and offset manually if needed\n // This is a workaround since the CLI might not support pagination\n return {\n content: [{type: \"text\", text: JSON.stringify(parsedOutput, null, 2)}],\n };\n }\n }\n \n console.error(`CLI error when listing vals, falling back to API: ${result.error}`);\n // Fall back to API on error\n } catch (error) {\n console.error(\"CLI error, falling back to API:\", getErrorMessage(error));\n // Fall back to API on error\n }\n }\n\n // API implementation (original code)\n try {\n const data = await callValTownApi(\n config,\n `/v2/me/vals?limit=${limit}&offset=${offset}`\n )\n return {\n content: [{type: \"text\", text: JSON.stringify(data, null, 2)}],\n }\n } catch (error) {\n console.error(error)\n return {\n content: [{type: \"text\", text: `Error listing your vals: ${getErrorMessage(error)}`}],\n isError: true,\n }\n }\n }\n )\n}",[],[278,340],[278],{"_105":348,"_95":96,"_157":349,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":350,"_101":32,"_165":166,"_167":-5,"_168":22,"_169":-5,"_170":351,"_99":352},"d73399dc-77b0-11f0-a503-0224a6c84d84","01989fb6-5dc4-7132-98d1-20e6b7231e93",["D",1755026185672],[],[96],{"_105":354,"_95":355,"_157":356,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":357,"_101":32,"_165":239,"_167":358,"_168":22,"_169":-5,"_170":359,"_99":360},"d7339ae0-77b0-11f0-a503-0224a6c84d84",".env.example","01989fb6-5dc4-7132-98d1-248f5d9e8542",["D",1755026185672],"# Val Town API token - get from https://www.val.town/settings\nVAL_TOWN_API_TOKEN=your_api_token_here\n",[],[355],{"_105":362,"_95":363,"_157":364,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":365,"_101":32,"_165":239,"_167":366,"_168":22,"_169":-5,"_170":367,"_99":368},"d7339bda-77b0-11f0-a503-0224a6c84d84",".gitignore","01989fb6-5dc4-7132-98d1-28c9eef58fbe",["D",1755026185672],"# Build outputs\n/build/\n\n# Environment variables\n.env\n\n# Deno\n.deno/\n\n# IDE-specific files\n.vscode/\n.idea/\n*.swp\n*.swo\n\n# OS-specific files\n.DS_Store\nThumbs.db\n",[],[363],{"_105":370,"_95":371,"_157":372,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":373,"_101":32,"_165":239,"_167":374,"_168":22,"_169":-5,"_170":375,"_99":376},"d7339cd4-77b0-11f0-a503-0224a6c84d84",".vtignore","01989fb6-5dc4-7132-98d1-2ca160b633e1",["D",1755026185672],".git\n.vscode\n.cursorrules\n.DS_Store\nnode_modules\nvendor\ntaskNotes/\nREADME.md\ndeno.lock\nCHANGELOG.md\nLICENSE\nmcp-documentation.md\nmcp.txt\n.github\ndocs\nbuild\ntaskNotes\nvalTownApi.json\nvaltownAPIv2.json\nval-town-mcp-server\n.claude\nmod.ts",[],[371],{"_105":378,"_95":379,"_157":380,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":381,"_101":32,"_165":239,"_167":382,"_168":22,"_169":-5,"_170":383,"_99":384},"d7339df6-77b0-11f0-a503-0224a6c84d84","CLAUDE.md","01989fb6-5dc4-7132-98d1-32f8e9bf095b",["D",1755026185672],"# CLAUDE.md\n\nThis file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.\n\n## About ValTown\n\nValTown is a serverless platform for running TypeScript/JavaScript code in the cloud. Key concepts for developers:\n\n- **Vals**: Code snippets that can be HTTP endpoints, scheduled functions, or scripts\n- **HTTP Vals**: Functions that handle HTTP requests (similar to serverless functions)\n- **Interval Vals**: Functions that run on a schedule (like cron jobs)\n- **Script Vals**: One-time or manually triggered functions\n- **Email Vals**: Functions that handle incoming emails\n- **API Integration**: Full REST API for programmatic access to create/update/delete vals\n- **Remote Execution**: Code runs in ValTown's sandboxed environment with built-in npm package support\n\n## Architecture Overview\n\nThis is a Model Context Protocol (MCP) server that provides AI assistants with access to ValTown's platform. The codebase supports two deployment modes:\n\n1. **Local Binary Mode** (`mod.ts`): Runs as a local MCP server using stdio transport\n2. **Remote HTTP Mode** (`index.http.ts`): Deployed on ValTown itself as an HTTP endpoint, accessible via streamable HTTP transport\n\nThe server is hosted remotely on ValTown at: `https://prashamtrivedi-valtown-mcp-server.val.run/mcp`\n\nKey files:\n- `mod.ts`: Local stdio MCP server entry point\n- `index.http.ts`: Remote HTTP MCP server entry point (deployed on ValTown)\n- `registerTools.ts`: Tool registration with CLI exclusion support for remote deployment\n- `config.ts`: Configuration loading with remote/local mode support\n\n## Commands\n\n- Run server locally: `deno task start`\n- Development mode: `deno task dev`\n- Build binaries: `deno task build`\n- Platform-specific builds: `deno task build:linux`, `deno task build:windows`, `deno task build:macos`\n\n## Style Guidelines\n\n- Use TypeScript strict mode with explicit typing for function parameters/returns\n- 2-space indentation, semicolons at line ends\n- Use camelCase for variables/functions, PascalCase for types/interfaces, UPPER_SNAKE_CASE for constants\n- Tool names use kebab-case (e.g., `get-val`)\n- Document all public functions with JSDoc\n- Wrap async operations in try/catch blocks and use getErrorMessage utility\n- Return consistent error responses with isError flag\n\n## Conventions\n\n- Follow [Conventional Commits](https://www.conventionalcommits.org/) format\n- Feature branches: `feature/description`, bug fixes: `fix/description`\n- Validate all user inputs and follow least privilege principle\n- Never commit API tokens or secrets\n- Each tool should have a clear purpose with descriptive parameters",[],[379],{"_105":386,"_95":387,"_157":388,"_159":-5,"_160":140,"_161":389,"_162":163,"_107":390,"_101":32,"_165":239,"_167":391,"_168":22,"_169":-5,"_170":392,"_99":393},"6676cd3a-77b1-11f0-b40f-0224a6c84d84","OAUTH_DEPLOYMENT.md","6676ce02-77b1-11f0-b40f-0224a6c84d84",4,["D",1755026426025],"# OAuth 2.1 Deployment Guide\n\nThis guide explains how to deploy the ValTown MCP Server with OAuth 2.1 support enabled.\n\n## Prerequisites\n\n1. ValTown account with API token\n2. RSA key pair for JWT signing\n3. Environment variables configured\n\n## Step 1: Generate RSA Key Pair\n\nGenerate the required RSA keys for JWT signing:\n\n```bash\n# Generate private key\nopenssl genrsa -out private.pem 2048\n\n# Generate public key \nopenssl rsa -in private.pem -pubout -out public.pem\n\n# Convert to single-line format for environment variables\nPRIVATE_KEY=$(awk 'NF {sub(/\\r/, \"\"); printf \"%s\\\\n\",$0;}' private.pem)\nPUBLIC_KEY=$(awk 'NF {sub(/\\r/, \"\"); printf \"%s\\\\n\",$0;}' public.pem)\n```\n\n## Step 2: Configure Environment Variables\n\nSet the following environment variables in your ValTown environment or .env file:\n\n```bash\n# OAuth Configuration\nOAUTH_ENABLED=true\nOAUTH_ISSUER_URL=https://prashamtrivedi-valtown-mcp-server.val.run\nOAUTH_JWT_PRIVATE_KEY=\"-----BEGIN PRIVATE KEY-----\\nMIIEvQIBADANBgkqhkiG9w0BAQEFAASCBKcwggSjAgEAAoIBAQC...\\n-----END PRIVATE KEY-----\"\nOAUTH_JWT_PUBLIC_KEY=\"-----BEGIN PUBLIC KEY-----\\nMIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEAtP...\\n-----END PUBLIC KEY-----\"\nOAUTH_ENCRYPTION_KEY=\"base64-encoded-32-byte-key\"\nOAUTH_ACCESS_TOKEN_TTL=3600\nOAUTH_REFRESH_TOKEN_TTL=2592000\n\n# ValTown Configuration (existing)\nVAL_TOWN_API_TOKEN=your_valtown_token_here\n```\n\n### Generating Encryption Key\n\n```bash\n# Generate 32-byte encryption key\nopenssl rand -base64 32\n```\n\n## Step 3: Deploy to ValTown\n\n1. Upload the updated code to your ValTown val\n2. Ensure all environment variables are set\n3. Test the deployment\n\n## Step 4: Verify OAuth Endpoints\n\nTest that all OAuth endpoints are working:\n\n```bash\n# Authorization Server Metadata\ncurl https://your-server.val.run/.well-known/oauth-authorization-server\n\n# Protected Resource Metadata \ncurl https://your-server.val.run/.well-known/oauth-protected-resource\n\n# JWKS Endpoint\ncurl https://your-server.val.run/.well-known/jwks.json\n\n# Client Registration\ncurl -X POST https://your-server.val.run/oauth/register \\\n -H \"Content-Type: application/json\" \\\n -d '{\n \"redirect_uris\": [\"https://client.example.com/callback\"],\n \"client_name\": \"Test Client\"\n }'\n```\n\n## Step 5: Test OAuth Flow\n\nRun the automated test suite:\n\n```bash\n# Set test environment\nexport TEST_SERVER_URL=https://your-server.val.run\nexport VAL_TOWN_API_TOKEN=your_token\n\n# Run OAuth compliance tests\ndeno run --allow-net --allow-env oauth-test.ts\n```\n\n## MCP Client Configuration\n\nFor MCP clients supporting OAuth, use these configuration values:\n\n### Claude Desktop Config\n```json\n{\n \"mcpServers\": {\n \"valtown-oauth\": {\n \"command\": \"npx\", \n \"args\": [\"@modelcontextprotocol/server-everything\"],\n \"transport\": {\n \"type\": \"http\",\n \"url\": \"https://your-server.val.run/mcp\",\n \"oauth\": {\n \"authorization_endpoint\": \"https://your-server.val.run/oauth/authorize\",\n \"token_endpoint\": \"https://your-server.val.run/oauth/token\", \n \"client_registration_endpoint\": \"https://your-server.val.run/oauth/register\",\n \"scopes\": [\"read\", \"write\", \"vals:read\", \"vals:write\"]\n }\n }\n }\n }\n}\n```\n\n### Manual OAuth Client Setup\n\n1. **Register Client:**\n ```bash\n curl -X POST https://your-server.val.run/oauth/register \\\n -H \"Content-Type: application/json\" \\\n -d '{\n \"redirect_uris\": [\"https://your-app.com/oauth/callback\"],\n \"client_name\": \"Your MCP Client\",\n \"scope\": \"read write vals:read vals:write\"\n }'\n ```\n\n2. **Store client_id and client_secret from response**\n\n3. **Authorization Request (PKCE):**\n ```bash\n # Generate PKCE parameters\n CODE_VERIFIER=$(openssl rand -base64 32 | tr -d \"=+/\" | cut -c1-43)\n CODE_CHALLENGE=$(echo -n $CODE_VERIFIER | openssl dgst -sha256 -binary | base64 | tr -d \"=+/\" | cut -c1-43)\n \n # Redirect user to:\n https://your-server.val.run/oauth/authorize?response_type=code&client_id=CLIENT_ID&redirect_uri=REDIRECT_URI&scope=read%20write&code_challenge=CODE_CHALLENGE&code_challenge_method=S256&state=RANDOM_STATE\n ```\n\n4. **Exchange Code for Token:**\n ```bash\n curl -X POST https://your-server.val.run/oauth/token \\\n -H \"Content-Type: application/json\" \\\n -d '{\n \"grant_type\": \"authorization_code\",\n \"client_id\": \"CLIENT_ID\",\n \"client_secret\": \"CLIENT_SECRET\",\n \"code\": \"AUTH_CODE\",\n \"redirect_uri\": \"REDIRECT_URI\",\n \"code_verifier\": \"CODE_VERIFIER\"\n }'\n ```\n\n5. **Use Access Token:**\n ```bash\n curl -X POST https://your-server.val.run/mcp \\\n -H \"Authorization: Bearer ACCESS_TOKEN\" \\\n -H \"Content-Type: application/json\" \\\n -d '{\"jsonrpc\": \"2.0\", \"id\": 1, \"method\": \"tools/list\"}'\n ```\n\n## Security Considerations\n\n1. **Environment Variables**: Store sensitive values securely\n2. **HTTPS Only**: Never use OAuth over HTTP in production \n3. **Key Management**: Rotate keys regularly\n4. **Scope Principle**: Grant minimal required scopes\n5. **Token Monitoring**: Monitor for suspicious activity\n\n## Troubleshooting\n\n### Common Issues\n\n1. **\"OAuth not enabled\" errors**: Check `OAUTH_ENABLED=true` is set\n2. **JWT signature failures**: Verify private/public key pair matches\n3. **Client registration fails**: Ensure redirect URIs use HTTPS\n4. **Token validation fails**: Check token hasn't expired\n5. **Scope errors**: Verify client has required scopes\n\n### Debug Mode\n\nEnable additional logging:\n```bash\nexport DEBUG=oauth:*\n```\n\n### Test OAuth Without MCP Client\n\nUse the test script to verify OAuth implementation:\n```bash\ndeno run --allow-net --allow-env oauth-test.ts\n```\n\n## Migration from Legacy Authentication\n\nThe system maintains backward compatibility:\n\n- **Existing clients** using `X-Val-Town-Token` continue to work\n- **New clients** can use OAuth for enhanced security\n- **Gradual migration** is supported\n\nTo migrate existing integrations:\n1. Deploy OAuth-enabled server\n2. Register OAuth clients\n3. Update client applications gradually\n4. Eventually deprecate legacy authentication\n\n## Support\n\nFor issues or questions:\n1. Check the test suite output for specific errors\n2. Verify environment variable configuration \n3. Review server logs for OAuth-related messages\n4. Consult the OAuth 2.1 specification for compliance details",[],[387],{"_105":395,"_95":396,"_157":397,"_159":-5,"_160":140,"_161":398,"_162":163,"_107":399,"_101":32,"_165":239,"_167":400,"_168":22,"_169":-5,"_170":401,"_99":402},"6672d93c-77b1-11f0-8de7-0224a6c84d84","complexity-assessment.md","6672d9dc-77b1-11f0-8de7-0224a6c84d84",3,["D",1755026426013],"# Complexity Assessment Report\n\n## Request Summary\n\nAdd OAuth 2.1 compliant authorization to the ValTown MCP server following official MCP authorization specifications. The current system has basic token-based auth via headers, but needs to be upgraded to full OAuth 2.1 compliance with metadata endpoints, dynamic client registration, and PKCE flow support.\n\n## Complexity Score: 4\n\n## Reasoning\n\n- **Cross-system OAuth Integration**: Requires implementing OAuth 2.1 server capabilities or integrating with external OAuth provider\n- **Multiple New Endpoints**: Need /.well-known/oauth-authorization-server, /.well-known/oauth-protected-resource, dynamic client registration endpoint\n- **Security Protocol Implementation**: PKCE flow, token validation, audience validation, secure token exchange\n- **Architecture Changes**: Current simple header-based auth needs to be replaced with full OAuth flow\n- **Third-party Integration Complexity**: ValTown doesn't have OAuth server, so need to bridge API token system with OAuth specs\n\n## Impact Analysis\n\n- **Backend Changes**: Major - Multiple new endpoints, OAuth flow implementation, metadata generation\n- **Frontend Changes**: None - This is pure backend OAuth server functionality \n- **Data Model Changes**: Yes - Client registration storage, token management, session handling\n- **API Changes**: New - Multiple new OAuth endpoints (.well-known, /oauth/authorize, /oauth/token, etc.)\n- **Breaking Changes**: No - Can maintain backward compatibility with current header-based auth\n\n## Recommended Approach\n\n### For Score 4 (Complex):\n\nExecute: `/complex-pipeline oauth-authorization`\n\n- Comprehensive planning with OAuth 2.1 specification review\n- Phased implementation:\n 1. OAuth metadata endpoints\n 2. Dynamic client registration \n 3. PKCE authorization flow\n 4. Token validation and audience checks\n 5. Integration with existing ValTown API token system\n- Multiple validation gates for OAuth compliance testing\n\n## Risk Assessment\n\n- **Technical Risk**: High - OAuth 2.1 compliance is complex, security-critical\n- **Business Risk**: Medium - Authorization changes could break existing integrations\n- **Rollback Complexity**: Moderate - Can maintain dual auth methods during transition\n\n## Key Implementation Challenges\n\n1. **ValTown API Token Bridge**: Need to map ValTown's simple API tokens to OAuth token format\n2. **OAuth Server Implementation**: Either implement minimal OAuth server or use library\n3. **PKCE Flow**: Implement secure code challenge/verifier mechanism\n4. **Metadata Generation**: Dynamic generation of OAuth server and protected resource metadata\n5. **Client Registration**: Storage and management of dynamically registered clients\n\n## Next Steps\n\n1. Review this assessment\n2. Approve the complexity score \n3. Execute `/complex-pipeline oauth-authorization`\n4. Backend OAuth implementation MUST be completed and validated before any client-side integration testing",[],[396],{"_105":404,"_95":405,"_157":406,"_159":-5,"_160":140,"_161":120,"_162":22,"_107":407,"_101":32,"_165":190,"_167":408,"_168":22,"_169":-5,"_170":409,"_99":410},"d733a1b6-77b0-11f0-a503-0224a6c84d84","config.ts","01989fb6-5dc4-7132-98d1-370e684e69f7",["D",1755026185672],"import {config as dotenvConfig} from \"https://deno.land/x/dotenv@v3.2.2/mod.ts\"\nimport * as path from \"jsr:@std/path\"\nimport {getCliAvailability} from \"./lib/vtCli.ts\"\nexport async function loadConfig(remoteMode = false) {\n if (remoteMode) {\n // For remote: expect token in request headers, use local prompt file\n return {\n apiToken: null, // Will be set from headers\n apiBase: \"https://api.val.town\",\n cli: {\n preferCli: false,\n path: \"vt\",\n },\n prompts: {\n valleyPath: `${import.meta.dirname}/prompts/valley.txt`, // Use local file in deployment\n defaultValleyPath: `${import.meta.dirname}/prompts/valley.txt`,\n },\n };\n }\n\n // Existing local config logic\n dotenvConfig({ export: true });\n\n const API_TOKEN = Deno.env.get(\"VAL_TOWN_API_TOKEN\");\n if (!API_TOKEN) {\n console.error(\"Error: VAL_TOWN_API_TOKEN environment variable is required\");\n Deno.exit(1);\n }\n\n // CLI preference configuration\n const PREFER_CLI = Deno.env.get(\"VAL_TOWN_PREFER_CLI\") !== \"false\"; // Default to true\n const CLI_PATH = Deno.env.get(\"VAL_TOWN_CLI_PATH\") || \"vt\"; // Default to \"vt\" in PATH\n // Prompt file paths (can be absolute or relative to CWD)\n const VALLEY_PROMPT_PATH = Deno.env.get(\"VALLEY_PROMPT_PATH\");\n const useCliIfAvailable = PREFER_CLI ?? false;\n const cliAvailable = useCliIfAvailable && await getCliAvailability();\n const defaultPromptPath = `${import.meta.dirname}/prompts/${\n cliAvailable ? \"valley.txt\" : \"valley_local.txt\"\n }`;\n\n // Resolve paths relative to the current working directory if they're not absolute\n const resolvedValleyPath = VALLEY_PROMPT_PATH\n ? (path.isAbsolute(VALLEY_PROMPT_PATH)\n ? VALLEY_PROMPT_PATH\n : defaultPromptPath)\n : undefined;\n return {\n apiToken: API_TOKEN,\n apiBase: \"https://api.val.town\",\n cli: {\n preferCli: PREFER_CLI,\n path: CLI_PATH,\n },\n prompts: {\n valleyPath: resolvedValleyPath,\n // Default paths included with the binary\n defaultValleyPath: defaultPromptPath,\n },\n };\n}",[],[405],{"_105":412,"_95":413,"_157":414,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":415,"_101":32,"_165":239,"_167":416,"_168":22,"_169":-5,"_170":417,"_99":418},"d733a3d2-77b0-11f0-a503-0224a6c84d84","deno.json","01989fb6-5dc4-7132-98d1-3ad0b5afd689",["D",1755026185672],"{\n \"tasks\": {\n \"start\": \"deno run --allow-net --allow-env --allow-read mod.ts\",\n \"dev\": \"deno run --watch --allow-net --allow-env --allow-read mod.ts\",\n \"build\": \"deno task build:linux && deno task build:windows && deno task build:macos\",\n \"build:linux\": \"deno compile --output build/valtown-mcp-linux --include prompts/valley.txt --include prompts/valley_local.txt --target x86_64-unknown-linux-gnu --allow-net --allow-env --allow-read mod.ts\",\n \"build:windows\": \"deno compile --output build/valtown-mcp-windows.exe --include prompts/valley.txt --include prompts/valley_local.txt --target x86_64-pc-windows-msvc --allow-net --allow-env --allow-read mod.ts\",\n \"build:macos\": \"deno compile --output build/valtown-mcp-macos --include prompts/valley.txt --include prompts/valley_local.txt --target x86_64-apple-darwin --allow-net --allow-env --allow-read mod.ts\"\n }\n}",[],[413],{"_105":420,"_95":100,"_157":421,"_159":-5,"_160":140,"_161":120,"_162":22,"_107":422,"_101":32,"_165":423,"_167":424,"_168":22,"_169":-5,"_170":425,"_99":426},"d733a4d6-77b0-11f0-a503-0224a6c84d84","01989fb6-5dc4-7132-98d1-3fa36270ccf6",["D",1755026185672],"http","/**\n * ValTown Remote HTTP MCP Server\n * \n * This is the HTTP entry point for the ValTown MCP Server when deployed\n * on ValTown itself. It excludes CLI-dependent functionality and uses\n * header-based authentication with Hono and StreamableHTTPServerTransport.\n */\n\nimport {McpServer} from \"npm:@modelcontextprotocol/sdk/server/mcp.js\"\nimport {StreamableHTTPServerTransport} from \"npm:@modelcontextprotocol/sdk/server/streamableHttp.js\"\nimport {Hono} from \"npm:hono\"\nimport {cors} from \"npm:hono/cors\"\nimport {toFetchResponse, toReqRes} from \"npm:fetch-to-node\"\nimport {registerPromptsTools, registerTools} from \"./registerTools.ts\"\nimport {loadConfig} from \"./config.ts\"\n\nconst app = new Hono()\n\n// Add CORS middleware\napp.use(\"/*\", cors({\n origin: \"*\",\n allowMethods: [\"GET\", \"POST\", \"OPTIONS\"],\n allowHeaders: [\"Content-Type\", \"Authorization\", \"Mcp-Session-Id\", \"X-Val-Town-Token\"],\n}))\n\n// Handle GET requests - Server info\napp.get(\"/\", (c) => {\n const responseData = {\n name: \"ValTown MCP Server\",\n version: \"1.0.0\",\n description: \"Model Context Protocol server for ValTown integration\",\n status: \"active\",\n message: \"This is an MCP server. Configuration endpoint coming soon...\",\n endpoints: {\n mcp: \"POST / - MCP protocol requests (requires X-Val-Town-Token header)\",\n info: \"GET / - This endpoint\"\n },\n documentation: \"https://github.com/PrashamTrivedi/valtown-mcp-server\"\n }\n\n return c.json(responseData)\n})\n\n\n// Handle MCP requests\napp.post(\"/mcp\", async (c) => {\n try {\n // Extract API token from headers\n const apiToken = c.req.header(\"X-Val-Town-Token\") ||\n c.req.header(\"Authorization\")?.replace(\"Bearer \", \"\")\n\n if (!apiToken) {\n return c.json({\n jsonrpc: \"2.0\",\n error: {code: -32000, message: \"Missing API token in X-Val-Town-Token header or Authorization header\"},\n id: null\n }, 401)\n }\n\n // Load remote configuration\n const config = await loadConfig(true)\n config.apiToken = apiToken\n // Convert Hono request to Node.js-style req/res\n const {req, res} = toReqRes(c.req.raw)\n\n // Create streamable HTTP transport\n const transport = new StreamableHTTPServerTransport({\n sessionIdGenerator: undefined,\n })\n\n // Create MCP server instance\n const server = new McpServer({\n name: \"val-town-remote\",\n version: \"1.0.0\",\n })\n\n // Register tools (exclude CLI-dependent tools)\n registerTools(server, config, {excludeCli: true})\n registerPromptsTools(server, config)\n\n // Connect server to transport\n await server.connect(transport)\n\n // Handle the MCP request using streamable transport\n let requestBody\n try {\n requestBody = await c.req.json()\n } catch (error) {\n if (error instanceof SyntaxError) {\n return c.json({\n jsonrpc: \"2.0\",\n error: {code: -32700, message: \"Parse error\"},\n id: null\n }, 400)\n }\n throw error\n }\n\n await transport.handleRequest(req, res, requestBody)\n\n // Clean up on close\n res.on('close', () => {\n console.log('closing request')\n transport.close()\n server.close()\n })\n\n // Convert Node.js response back to Fetch response\n return toFetchResponse(res)\n\n } catch (error) {\n console.error(\"Error handling MCP request:\", error)\n\n return c.json({\n jsonrpc: \"2.0\",\n error: {\n code: -32603,\n message: \"Internal server error\"\n },\n id: null,\n }, 500)\n }\n})\n\napp.get(\"/mcp\", (c) => {\n console.log(\"Received GET MCP request - not supported in stateless mode\")\n return c.json({\n jsonrpc: \"2.0\",\n id: null,\n }, 200)\n})\n\nexport default app.fetch",[],[100],{"_105":428,"_95":429,"_157":430,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":431,"_101":32,"_165":190,"_167":432,"_168":22,"_169":-5,"_170":433,"_99":434},"d733a8a0-77b0-11f0-a503-0224a6c84d84","registerTools.ts","01989fb6-5dc4-7132-98d1-40b6ed0f0638",["D",1755026185672],"// registerTools.ts\nimport {McpServer} from \"npm:@modelcontextprotocol/sdk/server/mcp.js\"\nimport {Config} from \"./lib/types.ts\"\nimport {registerUserTools} from \"./tools/userTools.ts\"\nimport {registerValTools} from \"./tools/valsTools.ts\"\nimport {registerBranchTools} from \"./tools/branchTools.ts\"\nimport {registerFileTools} from \"./tools/fileTools.ts\"\nimport {registerSqliteTools} from \"./tools/sqliteTools.ts\"\nimport {registerBlobTools} from \"./tools/blobTools.ts\"\nimport {registerCliTools} from \"./tools/cliTools.ts\"\nimport {registerPrompts} from \"./prompts/promptsTools.ts\"\n\nexport function registerTools(\n server: McpServer,\n config: Config,\n options?: { excludeCli?: boolean },\n) {\n registerUserTools(server, config);\n registerValTools(server, config); // Now handles vals (formerly projects)\n registerBranchTools(server, config);\n registerFileTools(server, config);\n registerSqliteTools(server, config);\n registerBlobTools(server, config);\n\n // Conditionally register CLI tools\n if (!options?.excludeCli) {\n registerCliTools(server, config); // Register CLI-specific tools\n }\n}\n\nexport function registerPromptsTools(server: McpServer, config: Config) {\n registerPrompts(server, config)\n}",[],[429],"files",[154,173,179,185,195,203,211,219,228,234,243,251,259,267,276,282,290,298,306,314,322,330,338,347,353,361,369,377,385,394,403,411,419,427],"counts",{"_439":32,"_440":182,"_441":182,"_442":32},"pulls","branches","forks","envVars","contents",[],{"_105":420,"_95":100,"_157":421,"_159":-5,"_160":140,"_161":120,"_162":22,"_107":446,"_101":32,"_165":423,"_167":424,"_168":22,"_169":-5,"_99":447},["D",1755026185672],[100],"upstreamProject","upstreamBranch",{"_105":143,"_107":451,"_452":453,"_95":43,"_141":106,"_101":145,"_142":-5,"_144":-5,"_454":455,"_109":-5,"_151":22,"_111":112,"_150":94,"_16":121,"_149":96},["D",1749829075618],"updatedAt",["D",1751349617129],"lockfileKey","1751349622569-esm.town_v_prashamtrivedi_val-town-http-mcp-server@57-main","emptyEnvVars","isNew","flags",[],"routes/_app.x.$handle.$name.branch.$branch.version.$version.code.$",{"_93":94,"_95":96,"_97":98,"_101":102,"_99":100,"_103":462,"_152":467,"_435":570,"_239":571,"_443":573,"_574":575,"_578":163,"_579":-5,"_580":581,"_582":32,"_583":584},{"_105":106,"_95":96,"_107":463,"_109":-5,"_110":22,"_111":112,"_113":72,"_114":-5,"_14":115,"_116":-5,"_117":22,"_118":-5,"_119":120,"_16":121,"_122":123,"_124":123,"_125":126,"_127":126,"_128":126,"_129":123,"_130":123,"_23":464,"_134":465,"_137":466,"_151":22},["D",1749829075618],{"_93":94,"_132":133},{"_93":94,"_136":133},{"_101":139,"_95":98,"_105":140,"_141":106,"_142":143,"_144":145,"_146":112,"_147":106,"_148":121,"_149":96,"_150":94},[468,471,492,510,537,540,543,546,549,552,555,558,561,564,567],{"_105":155,"_95":156,"_157":158,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":164,"_101":32,"_165":166,"_167":-5,"_168":22,"_169":-5,"_170":469,"_99":470},[],[156],{"_105":174,"_95":175,"_157":176,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":177,"_101":32,"_165":166,"_167":-5,"_168":22,"_169":-5,"_170":472,"_99":491},[473,479,482,485,488],{"_105":180,"_95":98,"_157":181,"_159":176,"_160":140,"_161":182,"_162":163,"_107":183,"_101":32,"_165":166,"_167":-5,"_168":22,"_169":-5,"_170":474,"_99":478},[475],{"_105":186,"_95":187,"_157":188,"_159":181,"_160":140,"_161":102,"_162":22,"_107":189,"_101":32,"_165":190,"_167":191,"_168":22,"_169":-5,"_170":476,"_99":477},[],[175,98,187],[175,98],{"_105":196,"_95":197,"_157":198,"_159":176,"_160":140,"_161":120,"_162":163,"_107":199,"_101":32,"_165":190,"_167":200,"_168":22,"_169":-5,"_170":480,"_99":481},[],[175,197],{"_105":204,"_95":205,"_157":206,"_159":176,"_160":140,"_161":120,"_162":163,"_107":207,"_101":32,"_165":190,"_167":208,"_168":22,"_169":-5,"_170":483,"_99":484},[],[175,205],{"_105":212,"_95":213,"_157":214,"_159":176,"_160":140,"_161":120,"_162":22,"_107":215,"_101":32,"_165":190,"_167":216,"_168":22,"_169":-5,"_170":486,"_99":487},[],[175,213],{"_105":220,"_95":221,"_157":222,"_159":176,"_160":140,"_161":120,"_162":163,"_107":223,"_101":32,"_165":190,"_167":224,"_168":22,"_169":-5,"_170":489,"_99":490},[],[175,221],[175],{"_105":229,"_95":230,"_157":231,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":232,"_101":32,"_165":166,"_167":-5,"_168":22,"_169":-5,"_170":493,"_99":509},[494,497,500,503,506],{"_105":235,"_95":236,"_157":237,"_159":231,"_160":140,"_161":120,"_162":163,"_107":238,"_101":32,"_165":239,"_167":240,"_168":22,"_169":-5,"_170":495,"_99":496},[],[230,236],{"_105":244,"_95":245,"_157":246,"_159":231,"_160":140,"_161":120,"_162":163,"_107":247,"_101":32,"_165":190,"_167":248,"_168":22,"_169":-5,"_170":498,"_99":499},[],[230,245],{"_105":252,"_95":253,"_157":254,"_159":231,"_160":140,"_161":120,"_162":163,"_107":255,"_101":32,"_165":239,"_167":256,"_168":22,"_169":-5,"_170":501,"_99":502},[],[230,253],{"_105":260,"_95":261,"_157":262,"_159":231,"_160":140,"_161":120,"_162":163,"_107":263,"_101":32,"_165":239,"_167":264,"_168":22,"_169":-5,"_170":504,"_99":505},[],[230,261],{"_105":268,"_95":269,"_157":270,"_159":231,"_160":140,"_161":120,"_162":163,"_107":271,"_101":32,"_165":239,"_167":272,"_168":22,"_169":-5,"_170":507,"_99":508},[],[230,269],[230],{"_105":277,"_95":278,"_157":279,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":280,"_101":32,"_165":166,"_167":-5,"_168":22,"_169":-5,"_170":511,"_99":536},[512,515,518,521,524,527,530,533],{"_105":283,"_95":284,"_157":285,"_159":279,"_160":140,"_161":120,"_162":163,"_107":286,"_101":32,"_165":190,"_167":287,"_168":22,"_169":-5,"_170":513,"_99":514},[],[278,284],{"_105":291,"_95":292,"_157":293,"_159":279,"_160":140,"_161":120,"_162":163,"_107":294,"_101":32,"_165":190,"_167":295,"_168":22,"_169":-5,"_170":516,"_99":517},[],[278,292],{"_105":299,"_95":300,"_157":301,"_159":279,"_160":140,"_161":120,"_162":163,"_107":302,"_101":32,"_165":190,"_167":303,"_168":22,"_169":-5,"_170":519,"_99":520},[],[278,300],{"_105":307,"_95":308,"_157":309,"_159":279,"_160":140,"_161":120,"_162":163,"_107":310,"_101":32,"_165":190,"_167":311,"_168":22,"_169":-5,"_170":522,"_99":523},[],[278,308],{"_105":315,"_95":316,"_157":317,"_159":279,"_160":140,"_161":120,"_162":163,"_107":318,"_101":32,"_165":190,"_167":319,"_168":22,"_169":-5,"_170":525,"_99":526},[],[278,316],{"_105":323,"_95":324,"_157":325,"_159":279,"_160":140,"_161":120,"_162":163,"_107":326,"_101":32,"_165":190,"_167":327,"_168":22,"_169":-5,"_170":528,"_99":529},[],[278,324],{"_105":331,"_95":332,"_157":333,"_159":279,"_160":140,"_161":120,"_162":163,"_107":334,"_101":32,"_165":190,"_167":335,"_168":22,"_169":-5,"_170":531,"_99":532},[],[278,332],{"_105":339,"_95":340,"_157":341,"_159":279,"_160":140,"_161":120,"_162":163,"_107":342,"_101":32,"_165":190,"_167":343,"_168":22,"_169":-5,"_170":534,"_99":535},[],[278,340],[278],{"_105":348,"_95":96,"_157":349,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":350,"_101":32,"_165":166,"_167":-5,"_168":22,"_169":-5,"_170":538,"_99":539},[],[96],{"_105":354,"_95":355,"_157":356,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":357,"_101":32,"_165":239,"_167":358,"_168":22,"_169":-5,"_170":541,"_99":542},[],[355],{"_105":362,"_95":363,"_157":364,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":365,"_101":32,"_165":239,"_167":366,"_168":22,"_169":-5,"_170":544,"_99":545},[],[363],{"_105":370,"_95":371,"_157":372,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":373,"_101":32,"_165":239,"_167":374,"_168":22,"_169":-5,"_170":547,"_99":548},[],[371],{"_105":378,"_95":379,"_157":380,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":381,"_101":32,"_165":239,"_167":382,"_168":22,"_169":-5,"_170":550,"_99":551},[],[379],{"_105":386,"_95":387,"_157":388,"_159":-5,"_160":140,"_161":389,"_162":163,"_107":390,"_101":32,"_165":239,"_167":391,"_168":22,"_169":-5,"_170":553,"_99":554},[],[387],{"_105":395,"_95":396,"_157":397,"_159":-5,"_160":140,"_161":398,"_162":163,"_107":399,"_101":32,"_165":239,"_167":400,"_168":22,"_169":-5,"_170":556,"_99":557},[],[396],{"_105":404,"_95":405,"_157":406,"_159":-5,"_160":140,"_161":120,"_162":22,"_107":407,"_101":32,"_165":190,"_167":408,"_168":22,"_169":-5,"_170":559,"_99":560},[],[405],{"_105":412,"_95":413,"_157":414,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":415,"_101":32,"_165":239,"_167":416,"_168":22,"_169":-5,"_170":562,"_99":563},[],[413],{"_105":420,"_95":100,"_157":421,"_159":-5,"_160":140,"_161":120,"_162":22,"_107":422,"_101":32,"_165":423,"_167":424,"_168":22,"_169":-5,"_170":565,"_99":566},[],[100],{"_105":428,"_95":429,"_157":430,"_159":-5,"_160":140,"_161":120,"_162":163,"_107":431,"_101":32,"_165":190,"_167":432,"_168":22,"_169":-5,"_170":568,"_99":569},[],[429],[468,471,473,475,479,482,485,488,492,494,497,500,503,506,510,512,515,518,521,524,527,530,533,537,540,543,546,549,552,555,558,561,564,567],{"_105":420,"_95":100,"_157":421,"_159":-5,"_160":140,"_161":120,"_162":22,"_107":446,"_101":32,"_165":423,"_167":424,"_168":22,"_169":-5,"_99":572},[100],[],"upstream",{"_105":143,"_107":576,"_452":577,"_95":43,"_141":106,"_101":145,"_142":-5,"_144":-5,"_454":455,"_109":-5,"_151":22,"_111":112,"_150":94,"_16":121,"_149":96},["D",1749829075618],["D",1751349617129],"showCompareCallout","pull","ahead",32,"behind","templates",[],"actionData","errors"]

Hi

Hi

Hello

Chat App

Hi

HTML content