AI-MEMORY AI-MEMORY
AI-MEMORY gives coding agents long-term, project-scoped memory that survives across sessions and across tools. Quit Claude Code mid-task, start OpenAI Codex, Gemini CLI, OpenCode, or Cursor in the same directory, and continue without re-explaining the architecture, the failed approaches, or the open questions.
Released v2.4.0 on September 21, 2026 with continued daily commits; the project surpassed 8,000 GitHub stars in September 2026 after trending on GitHub and gaining traction on X.
As of
- Cross-agent memory shared by Claude Code, Codex, Gemini CLI, OpenCode, Cursor, and others
- Automatic capture of decisions, failed approaches, and open questions via hooks
- MCP server for retrieving relevant memory on demand
- Session handoff so one agent can pick up exactly where another stopped
- Local-first storage per project directory with no cloud dependency
- Search and browse memory from the CLI
Developers who rotate between several coding agents on the same codebase and want to stop re-briefing each one on context, history, and dead ends.
Teams that need a hosted, multi-user shared memory service, or users who only ever run a single agent whose built-in memory is sufficient.
AI-MEMORY is free and MIT-licensed. Memory is stored locally per project; the only costs are the tokens consumed by the coding agents you already use, plus optional embedding models if semantic search is enabled.
AI-MEMORY solves a real, growing pain point for multi-agent workflows with a simple local-first design; it is young and moving fast, so expect frequent changes to the memory format and integrations.
Is AI-MEMORY free?
Yes - AI-MEMORY is Open Source. AI-MEMORY is free and MIT-licensed. Memory is stored locally per project; the only costs are the tokens consumed by the coding agents you already use, plus optional embedding models if semantic search is enabled.
Is AI-MEMORY open source?
Yes - AI-MEMORY is open source. AI-MEMORY is free and MIT-licensed. Memory is stored locally per project; the only costs are the tokens consumed by the coding agents you already use, plus optional embedding models if semantic search is enabled.
Who is AI-MEMORY best for?
Developers who rotate between several coding agents on the same codebase and want to stop re-briefing each one on context, history, and dead ends.
Who is AI-MEMORY not ideal for?
Teams that need a hosted, multi-user shared memory service, or users who only ever run a single agent whose built-in memory is sufficient.
What are the best AI-MEMORY alternatives?
The closest AI-MEMORY alternatives on ai.dosa.dev are Pieces for Developers, agenttrace, Vibe Kanban - all listed under Developer Productivity & Workflow.
Who makes AI-MEMORY?
AI-MEMORY is developed by AkitaOnRails. It is listed in the Developer Productivity & Workflow category on ai.dosa.dev.
Favorite this tool to revisit it later, or Zap it to contribute to the public vote count.
Content on this page is AI-generated. Please verify details with the vendor's website for accuracy.