os-environment-probe

To detect AI environments, this probe interviews you about your available tools and runs non-interactive probe commands to verify each claim. It reports pass or fail for every environment and persists the results.

Persisting AI environment details to context/memory/environment.md creates a single source of truth. Downstream skills read this profile to automatically select the appropriate delegation backend and cheapest brainstorm model.

You can verify installed AI tools by running lightweight probe commands against the environments you claim to have. The probe checks each tool non-interactively and reports a pass or fail status to update your profile.

You should run an environment probe on the first run after OS setup or whenever your AI environments change. Re-probing refreshes the results in context/memory/environment.md so delegation decisions remain accurate.

Yes, the probe checks access to multiple AI environments including Claude, Copilot CLI, Gemini CLI, and Cursor. It asks which tools you have, verifies each one with probe commands, and records the pass or fail results.

If a probe fails to find a claimed AI tool, it reports a fail status for that environment and updates the profile in context/memory/environment.md. Downstream skills then use the verified profile to select available delegation backends.