Skip to main content
Glama
alara7

claude-docs-mcp

by alara7

claude-docs-mcp

A local MCP server over the Claude documentation corpus. Built as hands-on preparation for the Claude Certified Architect – Foundations (CCA-F) exam, and as a portfolio artifact.

Status: specification and phase plan complete; implementation not started. See docs/implementation-plan.html for the current entry gate.

What it is

Three documentation sources are crawled offline into a Markdown corpus, chunked by heading, and indexed into SQLite FTS5. A separate stdio server exposes four read-only tools to Claude Code — search_docs, get_doc, get_doc_outline, list_sources — designed against a hard token budget so a single tool call never poisons the context window.

The server performs no network I/O at query time. Ingestion and serving are two binaries over one datastore, deliberately.

Related MCP server: MCP AI Docs

This repo contains the fetcher, not the documentation

The corpus is Anthropic's copyrighted documentation. corpus/ and data/ are gitignored and always will be. Run cdm fetch to build your own local copy.

Documents

File

What it is

docs/claude-docs-mcp-spec.html

The design specification — sources, schema, tool contracts, ADRs

docs/implementation-plan.html

Phase 0–7 plan with entry gates, exit criteria and handoff contracts

A full README — architecture diagram, transcripts, measured results and honest limitations — is Phase 7 of the plan.

Licence

MIT (see LICENSE, pending Phase 0).

Related MCP Connectors

Related MCP Servers

  • F
    license
    Not graded
    quality
    D
    maintenance
    A local-first MCP server that enables semantic search over PDF and DOCX documents using structure-aware parsing and vector storage. It allows users to query their local knowledge base through Claude Code without cloud dependencies or GPU requirements.
    -
  • F
    license
    A
    quality
    D
    maintenance
    An MCP server that provides indexed, searchable access to Anthropic Claude and Google Gemini documentation, with full-text search, page fetching, and section listing capabilities.
    4
    -