return-path-gate

Detect fire-and-forget dispatch patterns by auditing SKILL.md files to find where callers dispatch downstream skills without gating subsequent steps on callee verdicts. This skill classifies each chain as CLOSED or OPEN with severity tiers.

A fire-and-forget skill chain gap occurs when callers dispatch downstream skills but ignore their verdicts, leading to unverified outputs passing through undetected and breaking workflow logic in meta-harness plugin ecosystems.

Validate closure compliance by running a full chain audit across the plugins/ directory to scan SKILL.md files, classifying interconnected AI skills as CLOSED or OPEN with severity tiers before pipeline execution.

Fix non-compliant OPEN skill chains by applying targeted fix prescriptions generated during the audit, including special proposal-only flags for core harness skills that require deliberate human review.

Yes, use return-path-gate for pre-flight validation before running skill pipelines or submitting code changes to catch broken chain logic that would cause unverified AI outputs to propagate through your workflow.

Run a skill chain audit for AI workflow compliance when building interconnected AI skill workflows to ensure callers gate subsequent steps on callee verdicts, satisfying governance requirements for skill chain integrity.