Cookbook Bridge

Cookbook Bridge

**Cookbook (cookbook.team) is a shared workspace for a team and its AI agents.** Claude, ChatGPT/Codex, Gemini, Kimi and OpenClaw work on one task board, share one team memory, and return a receipt for every run: who ran it, what it did, how long it took, what it cost. This repository is the Bridge: the small program a member runs on their own machine so their own agents, on their own subscriptions, can do work for the team. No API keys, no token relay, nothing changes on your machine without your click.

• Product: https://cookbook.team
• What Cookbook is: https://cookbook.team/what-is-cookbook
• Docs: https://cookbook.team/docs
• Security model: https://cookbook.team/security
• npm: https://www.npmjs.com/package/cookbook-bridge
• The rest of Cookbook (the web app, the MCP server, the memory and receipts) is not in this repository. This mirror is published from the main repo on every release.

Changelog

**0.1.14** (2026-09-03)

• The AI-visibility probe rides the synthesis lane: a probe job asks one buyer prompt on your subscription, either with no tools (variant model) or with WebSearch only (variant search, passed as --tools WebSearch --allowedTools WebSearch because print mode refuses a tool nobody granted). Capped at 120s per prompt, logged as probe: # (), never with the prompt text.

**0.1.13** (2026-09-02)

• Pre-flight: the moment a grant this Bridge hosts becomes active, it runs env, the doctor, cli_versions and the shape of its own config (tokens stripped) once, locally and read-only, and posts the result as a host-initiated preflight call, so the visiting agent starts with what the machine already knows. Remembered per grant in bridge.state.json; a failed post is retried once per process.
• One-click plans: a plan call carries why and an ordered list of steps. The Bridge recomputes the plan hash before running (a step added after the click is refused as plan hash mismatch), runs each step