rtk-ai/icm

Permanent memory for AI agents. Single binary, zero dependencies, MCP native.

What it solves

ICM provides a shared, long-term memory system for AI coding agents. It prevents the need to repeatedly explain project conventions, decisions, and preferences every time a user starts a new session or switches between different AI tools (e.g., moving from Claude Code to Gemini CLI).

How it works

ICM stores memories in a single local SQLite file. It uses a hybrid recall system that combines three methods to find relevant information: keyword matching (BM25), semantic vector search (using sqlite-vec and an embedding model), and date-based filtering. It integrates with AI agents via hooks and an MCP (Model Context Protocol) server, allowing it to automatically inject relevant context into prompts and extract new facts from tool outputs.

Who it’s for

Developers who use multiple AI coding assistants and want a persistent, cross-tool memory of their project's technical decisions, architecture, and personal preferences.

Highlights

  • Cross-Tool Compatibility: Shared memory across up to 18 different agents and editors.
  • Local-First: Stores everything in one SQLite file with no external service required.
  • Efficient Storage: Does not require an LLM call to store or recall memories.
  • Hybrid Search: Combines keyword and semantic search for high retrieval accuracy.
  • Automatic Extraction: Can automatically extract facts from tool outputs to update memory.

Related

  • Project
    basicmachines-co/basic-memory

    A persistent, two-way memory system for AI assistants that uses shared Markdown files and the Model Context Protocol (MCP) to create a long-term knowledge graph.

  • Project
    atomicstrata/atomicmemory

    AtomicMemory is an open‑source, inspection‑ready memory service for LLM agents. It offers correction‑aware updates, model‑agnostic embeddings, and both managed‑cloud and self‑hosted Docker deployments. The project ships a Rust Core server, a TypeScript SDK, a CLI (`am`), adapters for LangChain, LangGraph, Vercel AI, etc., and plugins for various agent hosts. Quick start is a one‑line installer that creates a cloud profile; a `--local` flag runs the Core container on your machine. Memory can be ingested and queried via the CLI or programmatically via the SDK, and agents can be linked through an MCP server. Licensed under Apache 2.0.

  • Dispatch
    MCP-Memory: Fast Agent Memory with OKF and SQLite FTS5

    MCP-Memory is an OKF‑compliant server that gives AI agents persistent, searchable long‑term memory backed by SQLite FTS5, enabling sub‑20 ms lookups and human‑readable markdown storage.

  • Project
    MemTensor/MemOS

    MemOS 2.0 is a Memory Operating System for LLM‑based AI agents. It offers a unified API to store, retrieve, edit, and delete multi‑modal memories (text, images, tool traces) organized as a graph. The system can run as a hosted cloud service (API‑key access) or be self‑hosted via Docker (Neo4j + Qdrant) or a local SQLite plugin. Ready‑made plugins integrate MemOS with OpenClaw, Hermes, and DeepSeek Harness. Benchmarks show it outperforms existing commercial memory products. Licensed under Apache 2.0.

  • Project
    Lyellr88/marm-memory

    A persistent local memory server for AI agents using the Model Context Protocol (MCP), providing shared long-term memory and code-graph awareness across different AI clients.