headless-investigate-incident

Automate incident investigation planning by generating ranked hypotheses and parallel investigation tracks from organizational context like affected services and recent changes. It outputs structured data for AI review without gathering evidence or implementing fixes.

A parallel investigation track accelerates root cause analysis during incident response by structuring multiple independent investigative paths. The Skill generates these tracks alongside ranked hypotheses to guide autonomous testing and resolution efforts.

Generate ranked hypotheses by feeding system architecture, recent changes, and historical incident data into the planning process. The Skill produces three or more hypotheses complete with supporting evidence and testing steps.

No, you cannot use this skill to fix production issues. It strictly plans autonomous incident investigations by producing structured hypotheses and parallel tracks, and does not gather evidence or implement fixes.

Yes, the incident investigation outputs structured data specifically formatted for AI agent review. It uses the INCIDENT format containing distinct sections for HYPOTHESES and PARALLEL_TRACKS to facilitate automated downstream processing.

Provide context such as affected services, system architecture, recent changes, and historical incidents. The Skill applies this organizational context to alerts or symptom descriptions to generate the investigation plan.