marm-memory
Local-first 3-in-1 AI memory layer & MCP server for Claude Code, Codex, Grok, Gemini, VS Code and Cursor. Fuses session history, codebase indexing & concept gra
What is marm-memory?
Local-first 3-in-1 AI memory layer & MCP server for Claude Code, Codex, Grok, Gemini, VS Code and Cursor. Fuses session history, codebase indexing & concept graphs in SQLite. Enables zero-cloud, privacy-first context & instant recall, supports multi-agent swarms.
Overview of marm-memory
What it is
(from summary and README) marm-memory is a local-first 3-in-1 AI memory layer and MCP server. It fuses session history, codebase indexing and concept graphs in SQLite. It enables zero-cloud, privacy-first context and instant recall, and supports multi-agent swarms. It gives AI agents a permanent memory that persists across sessions and different AI assistants.
Key features
- Core Memory (8 tools): stores conversations, notes, notebook entries, and summaries so they stay searchable.
- Code Graph (6 tools): maps your repository so agents can find symbols, follow code paths, and understand the project without rereading it all.
- Concept Graph (2 tools): connects people, decisions, errors, and ideas from stored memories, with links back to relevant code when available.
- All 16 tools work over HTTP and STDIO.
- Bundled Console App provides a local control plane for memory, graphs, code context, distillation, runtime controls, and an integrated terminal.
- Supports multiple agent profiles (Claude, Codex, Grok, Gemini, Cursor, etc.) and multi-agent swarms via profiles like
swarm,swarm-max, andtrusted. - Performance layers: SQLite WAL mode, connection pooling, serialized write queue, HTTP rate-limit presets, FTS filter, semantic re-rank, and write-time deduplication.
Use cases
- Solo developer or researcher who wants persistent memory across chat sessions.
- Private local STDIO user who prefers not to use HTTP.
- Multiple agents sharing the same memory (e.g., Claude Code and Codex).
- Private high-throughput swarm for multi-agent collaboration.
- Trusted private lab or server environment.
Getting started
bash pip install marm-mcp-server marm-memory init --g-claude --g-codex --g-antigravity Then tell your AI agent: “Use the marm-init skill to set up MARM.” For manual setup, see the repository README for detailed client-specific commands.
License and pricing
Licensed under Apache-2.0. Open source and free.
Based on the project's public README and repository metadata.