preflight-risk-review

A preflight risk review converts ee preflight and tripwire JSON outputs into an evidence-backed brief with low, medium, high, blocked, or unknown risk levels. It enforces stop/go gates for destructive commands and migrations, preventing actions when evidence is missing or degraded.

A preflight risk assessment handles degraded or missing evidence by refusing to claim low risk from silence. It preserves degraded codes, redaction status, and provenance, enforcing strict gates that block the agent from proceeding without explicit confirmation and valid evidence bundles.

You need an evidence-bounded risk brief before coding-agent actions involving destructive commands, database migrations, production deploys, or dependency updates. It is required whenever you must verify operational safety using workspace-scoped ee JSON and exported tripwire evidence before executing changes.

Yes, you can use preflight tripwire outputs to generate must-verify checks, follow-up ee commands, and asks-now questions for deploys. The review process consumes explicit ee --json outputs and approved evidence bundles to plan operational verification and enforce stop/go gates.

Preflight risk reviews are limited by their strict dependence on consuming only explicit ee --json outputs and approved evidence bundles. They cannot claim low risk from silence, will block actions during degraded states, and require explicit confirmation for high-impact production deploys without overriding redaction or prompt-injection gates.

Yes, a preflight risk review preserves evidence provenance and IDs throughout redaction handling. It enforces redaction, prompt-injection, and degraded-state gates while emitting a schema-compliant ee.skill.preflight_risk_review.v1 brief that maintains full workspace-scoped evidence tracking.