mcp-typescript-docs

patpil-cloudflare-mcp/mcp-typescript-docs

3.2

If you are the rightful owner of mcp-typescript-docs and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.

Crawl4AI Documentation AI Search MCP provides intelligent search capabilities over the Crawl4AI library documentation using Cloudflare AutoRAG technology.

Tools
4
Resources
0
Prompts
0

MCP TypeScript SDK Documentation AI Search

Semantic search over Model Context Protocol (MCP) TypeScript SDK documentation using Cloudflare AutoRAG (AI Search).

What is This?

This MCP server provides intelligent search over MCP TypeScript SDK documentation from GitHub:

  • Semantic Search - Natural language queries over MCP SDK docs
  • AI-Generated Answers - LLM responses grounded in documentation
  • Dual Authentication - OAuth 2.1 and API key support

AI Search Instance: map-docs Domain: tsmcpdocs.wtyczki.ai

Use Cases:

  • Learn MCP SDK architecture and core concepts
  • Get TypeScript code examples for building MCP servers
  • Understand tool definitions, resource handlers, and prompt patterns
  • Troubleshoot MCP implementation issues

Available Tools

ToolDescription
search_mcp_docsSearch MCP TypeScript SDK documentation for SDK architecture, tool definitions, transport layers, and code examples

Production URLs

EndpointTransportDescription
https://tsmcpdocs.wtyczki.ai/sseServer-Sent EventsLegacy transport (Claude Desktop)
https://tsmcpdocs.wtyczki.ai/mcpStreamable HTTPModern transport (ChatGPT)

Features

  • AutoRAG (AI Search) Ready - Pre-configured Workers AI binding
  • Dual Transport Support - Both SSE (legacy) and Streamable HTTP
  • ChatGPT Compatible - Works with ChatGPT via /mcp endpoint
  • Claude Desktop Compatible - Works via /sse endpoint
  • Dual Authentication - OAuth 2.1 and API key support
  • WorkOS Magic Auth - Email + 6-digit code authentication

Testing

Pre-Deployment:

npx tsc --noEmit  # Must pass with zero errors

Post-Deployment (Cloudflare Workers AI Playground):

  1. Navigate to https://playground.ai.cloudflare.com/
  2. Set model to @cf/meta/llama-3.3-70b-instruct-fp8-fast
  3. Add MCP server: https://tsmcpdocs.wtyczki.ai/sse
  4. Complete OAuth flow
  5. Test search_mcp_docs tool with query: "How do I define tools in an MCP server?"

Project Structure

tsmcpdocs/
├── src/
│   ├── index.ts              # Entry point (dual transport)
│   ├── server.ts             # McpTypescriptDocsMCP with tools
│   ├── api-key-handler.ts    # API key authentication path
│   ├── authkit-handler.ts    # WorkOS OAuth handler
│   ├── types.ts              # Type definitions
│   ├── props.ts              # Auth context
│   ├── tokenUtils.ts         # Token management
│   ├── tokenConsumption.ts   # Token consumption logic
│   ├── apiKeys.ts            # API key validation
│   └── api-client.ts         # API client template
├── wrangler.jsonc            # Cloudflare config
├── package.json              # Dependencies
└── README.md                 # This file

Configuration

SettingValue
Worker Namemcp-typescript-docs
Class NameMcpTypescriptDocsMCP
AI Search Instancemap-docs
Domaintsmcpdocs.wtyczki.ai
Databasemcp-oauth (shared)

GitHub Integration

This repository uses Cloudflare Workers Builds for automatic deployments. Every push to main triggers an automatic deployment to tsmcpdocs.wtyczki.ai.