What it does
ThoughtCop reads what Claude Code and Codex are thinking while they work. You write any rule in plain English. Examples include skipping tests, touching secrets, working around a blocker, and claiming done without proof. ThoughtCop checks every thought and reply against your rules. It flashes a red bolt the moment a rule is broken. ThoughtCop needs no hook and no proxy. Nothing changes in the agent.
Capabilities
- Runs as a macOS menu-bar app.
- Reads the thoughts of both Claude Code and Codex while they work.
- Checks every thought and reply against the rules you define.
- Flashes a red bolt the moment a rule is broken.
- Needs no hook and no proxy, and changes nothing in the agent.
When to use it
- You want to know when your AI coding agent skips tests.
- You want to be alerted when your AI coding agent touches secrets or works around a blocker.
- You want to catch your AI coding agent claiming it is done without proof.
