docs-mcp
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@docs-mcpsearch docs for authentication"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
finance-docs-mcp
AI-searchable finance & investment knowledge base. MCP server + RAG chatbot for financial literacy education, built on BM25 + metadata filtering. Korean + English bilingual.
What This Does
You write MDX documentation with structured frontmatter. This system indexes those documents and exposes them as MCP tools that Claude Code (or any MCP client) can use to search, retrieve, and reason about your docs.
MDX Documents --> Indexer --> Search Index --> MCP Server --> AI Assistant
(BM25) (tools)Related MCP server: Alcove
Quick Start
# 1. Clone and install
git clone <repo-url> my-docs-mcp
cd my-docs-mcp
npm install
# 2. Configure
cp .env.example .env
# Edit .env: set OPENAI_API_KEY, DOCS_ROOT
# 3. Build the search index
npm run reindex
# 4. Start the MCP server
npm run startArchitecture
docs-mcp/
|
+-- content/
| +-- docs/ # Your MDX documentation
| | +-- domain-a/ # Organized by domain
| | +-- domain-b/
| +-- conventions/ # How-to-write-docs guides
|
+-- src/
| +-- indexer/ # MDX parser + BM25 index builder
| +-- retriever/ # Search engine (BM25 + metadata filters)
| +-- tools/ # MCP tool definitions
| +-- chat/ # RAG chatbot (OpenAI)
|
+-- analysis/
| +-- scripts/ # Analysis pipeline scripts
| +-- templates/ # MDX templates for generated docs
|
+-- data/
| +-- search-index.json # Built index (gitignored)
|
+-- dist/ # Compiled output (gitignored)Data Flow
MDX files (content/docs/)
|
v
[Indexer] -- parses frontmatter + content
|
v
search-index.json (BM25 terms + metadata)
|
v
[Retriever] -- query -> ranked results
|
v
[MCP Tools] -- exposed to AI assistants
|
v
[Chat] -- RAG: retriever results + OpenAI completionMCP Tools
Tool | Description |
| BM25 keyword search with optional category, domain, platform, and tag filters |
| Retrieve a single document by its path |
| List all categories with document counts |
| Find documents related to a given document (uses |
Example Queries
search_docs("user registration flow")
search_docs("traps", { category: "traps", domain: "user-management" })
get_document("example/sample-api")
list_categories()
find_related("example/index")Configuration
Environment Variables
Variable | Required | Default | Description |
| No |
| Path to MDX documents |
| For chat | -- | OpenAI API key for RAG chatbot |
| No |
| Server port |
| No |
| Log verbosity: debug, info, warn, error |
.mcp.json (Client Configuration)
Add to your project's .mcp.json to connect Claude Code:
{
"mcpServers": {
"project-docs": {
"command": "node",
"args": ["/path/to/docs-mcp/dist/server.js"],
"env": {
"DOCS_ROOT": "/path/to/docs-mcp/content/docs"
}
}
}
}Writing Documentation
Documents are MDX files with YAML frontmatter. Three fields are required:
---
title: "Document Title"
description: "One-line summary for search results"
category: endpoint
---See content/conventions/ for full documentation:
Frontmatter Schema -- All available fields
Category System -- When to use each category
FE/BE Separation -- Directory organization
Analysis Pipeline
The analysis/ directory contains scripts and templates for automated documentation generation. The pipeline:
Extract -- Parse source code (backend routes, frontend components) to identify endpoints and pages
Analyze -- Generate analysis documents (flows, traps, dependencies) using structured templates
Generate -- Output MDX files with correct frontmatter into
content/docs/Index -- Rebuild the search index with
npm run reindex
Templates in analysis/templates/ define the structure for each document category.
Deployment (Fly.io)
# First time
fly launch --name my-docs-mcp
# Deploy
fly deploy
# Set secrets
fly secrets set OPENAI_API_KEY=sk-...The server runs as a long-lived process. The search index is built at startup from the bundled MDX files.
Development
# Build
npm run build
# Run in development mode (auto-rebuild)
npm run dev
# Rebuild search index
npm run reindex
# Type check
npm run typecheckTech Stack
Runtime: Node.js 20+, TypeScript
Search: BM25 (no vector DB)
Protocol: MCP (Model Context Protocol)
Chat: OpenAI GPT-4 (RAG only, optional)
Deployment: Fly.io (or any Node.js host)
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- Alicense-qualityDmaintenanceAn MCP server implementation that provides tools for retrieving and processing documentation through vector search, enabling AI assistants to augment their responses with relevant documentation contextLast updated22265MIT
- AlicenseAqualityBmaintenanceMCP server that gives AI coding agents on-demand access to private project docs via BM25 ranked search. One setup for Claude Code, Cursor, Codex, Gemini CLI, and more. Docs stay private, never in public repos.Last updated1515Apache 2.0
- Alicense-qualityDmaintenanceGeneric MCP server that exposes Markdown documentation to LLMs, enabling them to search and answer questions about any software documentation.Last updatedMIT
- Alicense-qualityDmaintenanceTransforms Markdown documentation into an intelligent knowledge base with AI-powered search and Q\&A through an MCP server.Last updated259MIT
Related MCP Connectors
Local-first RAG engine with MCP server for AI agent integration.
An MCP server that gives your AI access to the source code and docs of all public github repos
Driflyte MCP server which lets AI assistants query topic-specific knowledge from web and GitHub.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/yundoun/finance-docs-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server