Show HN: Bourdon – one shared memory file for Claude Code, Codex, Cursor

Bourdon is a Show HN project that offers one shared memory file for Claude Code, Codex, and Cursor. The product is presented at bourdon.ai and was listed on Hacker News with 3 points and 0 comments at the time of this write-up.

The core idea is a single shared memory file that those three coding agents can use together. Instead of each tool keeping its own separate context or session memory, Bourdon is positioned as one common store that Claude Code, Codex, and Cursor can all read from and write to. No architecture diagrams, benchmarks, or implementation details were published in the listing beyond that product framing.

For engineers who switch among Claude Code, Codex, and Cursor in the same workflow, a single shared memory file is meant to cut down on re-explaining project state, preferences, and prior decisions when they change tools. The value depends on whether the file actually stays consistent across those clients and whether it fits how each agent already loads context.

The listing sits in the same lane as other agent-memory and multi-tool continuity efforts around Claude Code, Codex, and Cursor, where builders are trying to stop context from fragmenting across IDEs and CLI agents. At 3 points and no discussion yet on Hacker News, the project is still early in public visibility and has not drawn a competitive thread or user reports on the post.

Practical next steps are to try the shared memory file against a real multi-tool workflow on bourdon.ai, check whether Claude Code, Codex, and Cursor all honor the same file cleanly, and watch whether the HN thread or the site adds concrete setup steps, file format details, and limits as more people try it.