Local AI agent memory tool with MCP server and plugins
Your AI coding environment, restored on every machine. · Shared project memory for every coding agent, with encrypted Claude Code and Codex backup and restore in development preview. · From one machine's agent history to a restored coding environment. · See everything your agents knowin one local view. · Questions, answered. · Stop explaining yourselfto every new agent.
Docmancer is a specialized memory system designed for AI coding agents, enabling them to maintain a persistent, shared knowledge base. It runs locally, ensuring data privacy and control. The tool provides AI agents with consistent memory across sessions through a CLI interface, a Model Context Protocol (MCP) server for standardized communication, and installable plugins for extended functionality. It is built for developers and teams using AI assistants for software development, solving the problem of agents having fragmented, session-limited memory. This allows for more coherent long-term project work, better context retention, and reduced repetition in agent interactions.
pypi:docmancer
| Type | Agent |
| Section | RAG & vector databases |
| Pricing | has a free tier |
| Platform | Web + desktop |
| Systems | macos, cli, api, web |
| Hosting | local |
| Install | mcp |
| Protocols | mcp |
| Site language | en |
| GitHub | docmancer/docmancer |
| Rating | 0.00 (0 reviews) |
| Views | 8 |
| Launched | 2026-08-05 |