Memtomem

Long-term memory for AI coding agents, kept as Markdown you own. Hybrid BM25 + vector search over your notes and code via MCP — one local store shared across Cl…

LLM Mart 1 views 127 listing impressions

Memtomem is an open-source memory infrastructure for AI agents licensed under Apache 2.0. It provides both long-term memory (LTM) and short-term memory (STM) layers connected over MCP. The product stores memories as Markdown files the user owns, allowing work to be carried across different AI coding agents such as Claude Code, Codex, and Antigravity CLI.

Long-term memory combines BM25 keyword search with dense vector embeddings for searching notes, docs, and code. Memtomem is local-first and private, using SQLite storage and local ONNX inference without needing a GPU or cloud account. The optional STM proxy can attach relevant memories to MCP tool responses and compress responses before they reach the agent.

Summary drafted from the project's own website. Every sentence is backed by text on that page and was reviewed before publishing.

Related servers

Universal MCP Server with advanced AI memory capabilities and semantic search.

30 views

MCP server for MikroTik routers: firewall, NAT, routing, DHCP, DNS, WireGuard and more via SSH.

28 views

Let Codex orchestrate external coding agents through their native harnesses.

28 views

Control plane MCP for scoped recon, triage, and bounded proofs.

27 views