Resources

Everything needed to test an agent properly

Reference material, playbooks and the reasoning behind how this product decides what it is willing to claim.

Documentation

The connector contract, the evidence model and every check, with examples taken from what the engine parses.

Task-shaped walkthroughs: connecting an agent, triaging a failure, gating a release.

What building the detector taught us about testing agents that take actions.

Live checks against the API, the database and the run queue, performed when you load the page.