Browse
→ Code & Dev
→ Docmancer MCP Server
Docmancer MCP Server
Open source MCP server that discovers the memory, instructions, rules and session evidence that Claude Code, Codex, Cursor, Gemini, OpenCode, Cline, Windsurf and other AI coding agents already wrote on your machine. Combines them into one canonical Markdown tree and serves grounded, cited recall through tools like ask_memory and canonical_memory. Uses SQLite FTS5 for lexical search and dense vectors from a vendored static model with Reciprocal Rank Fusion for result merging. Local-first with no external API dependencies.
MCP unverified
Integration
| Transport | stdio |
| Auth | none |
| Endpoint | https://www.docmancer.dev |
| Install | npx docmancer mcp install claude-code |
Use Cases
| 01 | Unify and search all the memory files, instructions and rules that Claude Code, Codex, Cursor and other coding agents have created across your machine through a single MCP server instead of manually tracking scattered CLAUDE.md and memory files |
| 02 | Enable AI coding agents to recall previous decisions, patterns and project context through Docmancer MCP using semantic and lexical hybrid search with cited sources, reducing repeated explanations and context loss across sessions |
| 03 | Build a local-first knowledge base from existing agent memory that any MCP-compatible tool can query, with no external API dependencies and full control over what data stays on your machine |
Tags
memory developer-tools documentation search local-first coding-agents knowledge-management
Machine-readable: /api/servers.json
· JSON-LD schema embedded in <head>