Ai-memory

Ai-memory is persistent, shared long-term memory for AI coding agents such as Claude Code, Codex, Cursor, and OpenCode.

  • AI
  • Productivity
  • Tools & Utilities
Ai-memory is persistent, shared long-term memory for AI coding agents such as Claude Code, Codex, Cursor, and OpenCode. It records every agent session, builds a searchable knowledge base of what your agents have learned, and passes that context via a Model Context Protocol (MCP) server when you switch agents or start a new session. Designed for homelabbers who use multiple AI coding tools and want to stop re-explaining project architecture and decisions in every new session.

Platform availability

Available on 1 of 10 platforms

  • Unraid
  • TrueNAS (not listed)
  • Umbrel (not listed)
  • ZimaOS (not listed)
  • Proxmox (not listed)
  • Helm (not listed)
  • Mac (not listed)
  • Windows (not listed)
  • Linux (not listed)
  • Any server (not listed)

Health score?A 0-100 score computed nightly from four factors: maintained (40%: recent commits and steady activity), popular (25%: GitHub stars, log scale), easy to install (20%: platforms, install notes, setup guides, Docker image) and light to run (15%: minimum RAM, ARM support). When the source code has been checked, a fifth factor, security check (20%: automated scan for committed secrets, unsafe desktop-app settings, known-vulnerable dependencies and project hygiene; any serious finding caps it at 20), is added. Factors without data are left out and the rest are rescaled.

65/100 Good

Maintained
100last commit 2 days ago
Popular
8416,154 GitHub stars
Easy to install
351 install platform, install notes, Docker image
Light to run
No data
Security check
08 serious findings, 6 minor, CI, tests

Based on 4 of 5 factors.

Availability: platform listings confirmed today

GitHub stars
16,154
Open issues
24
Last commit
2026-11-19
Latest release
v3.5.1
License
MIT
Activity
steady

Checked 2 days ago - source: GitHub (akitaonrails/ai-memory)

Resources & Compatibility

ARM (e.g. Raspberry Pi) is supported.

Checked 2 days ago - source: Docker Hub image tags and published docs

First-install notes

  • Container & MCP setup Runs as a single Docker container with a built-in SQLite database — no external database or API keys required (an LLM provider is optional and improves summaries). Connect MCP clients to the server at port 49374/mcp, sending the Auth token as a Bearer token. On a client machine, install the CLI and run: ai-memory install-mcp --client <client-name> --auth-token YOUR_TOKEN. The Allowed Hosts setting source
  • Web console (HTTPS required) The built-in web console (/web) requires HTTPS. Place a reverse proxy (Traefik, Caddy, Nginx, etc.) in front of the container, enable the Secure cookie option, then set a Root username and an initial root password. Do not set a root password without HTTPS. source

Fetched 2 days ago - each note links to its source

Alternatives

No alternatives collected yet.

Common questions

How do MCP clients authenticate to ai-memory?

Connect them to the server's /mcp endpoint on port 49374 and send the auth token as a Bearer token. On a machine with the CLI, use ai-memory install-mcp --client claude-code --auth-token YOUR_TOKEN.

Why can ai-memory reject a client request?

The allowed hosts configuration must list every hostname or IP clients use to reach the server. Requests with any other Host header are refused.

How do I expose the web console securely?

The /web console needs HTTPS, so put a reverse proxy in front and set Secure cookie to true.

Are multi-user and audit features locked behind a paid plan?

No, multi-user auth, per-person attribution, and an audit log are built in and are not a paid tier.

Answers sourced from github.com

Community

Collected 2 days ago - public community threads