harness-governance

To enforce human approval for sensitive paths, classify the proposed action into a specific trust tier using governance gates. Higher-risk work, such as remote or production changes, then mandates appropriate authorization and sign-off before committing.

Trust tiers in AI governance classify actions into six levels ranging from read-only to remote production. This tiered risk mapping ensures that higher-risk development work requires explicit authorization and appropriate human sign-off.

To ensure documentation artifacts are included, apply companion rule requirements that map trigger changes to required satisfier documents. This governance mechanism mandates all companion documentation exists in the same PR or commit.

The best way to validate lifecycle readiness is to execute the expected validator chain and check exit-contract behavior. This verifies whether your project satisfies the Bootstrap Complete and Harness Ready lifecycle stages.

Update harness entrypoints and governance-critical files when changes impact operational governance entrypoints or sensitive path coverage. Validate skill content and documentation drift to ensure lifecycle readiness before opening a PR.

Yes, you can use CI validators to check compliance by specifying the expected validator chain and exit-contract behavior. This automates the validation of trigger-to-satisfier companion rules and lifecycle readiness checks.