Governance intensity
TrackeyAI v2.1.0.0.1 supports Low, Medium, and High (Beta) governance intensity. Select Intensity from the desktop dashboard or New Governed Run workflow. This setting controls governance and evidence-checking expectations, not which AI model is used or how intelligent it is.
Which level should I use?
| Setting | Intended workflow |
|---|---|
| Low | Lighter-weight governance and fewer check-ins where the run's Execution Contract explicitly permits it. Unperformed verification remains an assurance gap, never a verified result. |
| Medium | Balanced assurance with evidence at required checkpoints and substantive repairs before completion. The default governance profile. |
| High (Beta) | A stricter, checkpoint-driven workflow with deeper validation and evidence obligations. Intended for work requiring closer oversight. |
How governance works
TrackeyAI evaluates the work against the governed run's objective, requirements, allowed scope, required evidence, and acceptance criteria. The authoritative decisions are PASS, FIX, and STOP. An AI agent's assertion of completion is not a substitute for a TrackeyAI decision.
The chosen governance profile is recorded in the run's Execution Contract. Changing the live intensity setting does not retroactively rewrite existing run contracts. Some Low-intensity efficiencies apply only when explicitly enabled in a new contract.
A PASS at an intermediate checkpoint does not necessarily mean a whole run is closed. Check the run's final disposition and whether run_completion_closed is true.
For run statuses, reports, and next actions, see Governed runs and statuses. For agent integration, see Cursor MCP integration.