A local setup where several AI tools share one memory
Using Claude Code, Codex, Grok, and Hermes side by side can create four separate memories for the same work. This setup uses one as the real source of truth, so the AI tools read the same Markdown notes and write lasting facts back into the same place. A middle layer called gbrain acts as the shared brain.
It indexes the into PGLite, uses nomic-embed-text through Ollama to make the notes easier to search, and builds a . Claude Code, Codex, Grok, and Hermes all connect to the same HTTP MCP server, so each tool queries the same memory. The server on 127.0.0.1 and uses bearer-token auth.
There is no external API, so the data stays on the machine and each query costs nothing.
Key points
- One is used as the shared source of truth for multiple AI tools.
- Claude Code, Codex, Grok, and Hermes all query the same HTTP MCP server.
- gbrain indexes the Markdown notes into PGLite and builds a .
- nomic-embed-text through Ollama, with no external API calls.
- The setup keeps data on the device and has no per-query cost.