How to install ai-memory so Claude Code and Codex share one memory

ai-memory is an open-source tool that lets several coding agents, including Claude Code and Codex, share one memory. Lifecycle hooks quietly record prompts, tool calls and session boundaries, and at session end they're consolidated into per-project markdown wiki pages. Quit Claude Code mid-task, open Codex in the same folder, and it picks up where you left off and what failed. The default path makes zero LLM calls, so no API key is needed. Version 3.1.0 shipped on October 16 and the repo hit GitHub trending.

1. Install the binary and run setup. It creates the data directory and token, starts a per-user service (launchd or systemd --user), and asks which detected agents to wire with MCP, hooks and skills.

curl -fsSL https://github.com/akitaonrails/ai-memory/releases/latest/download/install.sh | sh
ai-memory setup

1. Claude Code and Codex can also be wired as marketplace plugins (the binary is still required, 3.0 or later). Codex then asks you to trust the hooks in /hooks.

claude plugin marketplace add akitaonrails/ai-memory && claude plugin install ai-memory@ai-memory
codex plugin marketplace add akitaonrails/ai-memory && codex plugin add ai-memory@ai-memory

1. Restart your agents and ask one "What do you remember about this project?" to confirm it answers through ai-memory.

2. To hand work from Claude Code to Codex, launch the agents with ai-memory run. Since 3.1.0, a switch in the same checkout also carries a bounded projection of the previous agent's session history.

ai-memory run claude
ai-memory run codex

1. New in 3.1.0, --global searches the wiki pages of every project from the terminal.

ai-memory search --global "auth token expiry"

macOS, Linux and WSL2 are supported; native Windows is experimental. Memory is git-backed markdown, so you can grep it or edit it by hand.