Lists multiple specific concrete actions: triage the situation, decide between stabilize-and-mitigate vs rollback vs drift reconciliation, route to the right mutation skill with scope and --change-desc, and drive post-incident verification + close-out.

Clearly answers both 'what' (triage, decide between mitigation strategies, route to mutation skills, drive post-incident verification) and 'when' (explicit 'Use when' clause with extensive trigger phrases). Also includes explicit 'do not use when' guidance for disambiguation.

Excellent coverage of natural phrases users would actually say during an incident: 'we have an outage', 'prod is crashing', 'page me through this', 'what do I do first?', 'mitigate or roll back?', 'production is down', 'something's broken in staging', 'post-incident cleanup'. These are highly realistic trigger phrases.

Highly distinctive with a clear niche (production incidents/outages). Explicitly disambiguates from related skills by naming them ('promote-release' for planned releases, 'cub-mutate' for single-Unit edits), which strongly reduces conflict risk.

The skill is thorough and mostly earns its length given the complexity of incident orchestration, but some sections are slightly verbose — e.g., the 'Principle' section, the 'Do not load for' list (which duplicates frontmatter), and some explanatory asides like 'This costs nothing in the moment and saves hours in the postmortem' that Claude doesn't need to be told. Overall reasonably efficient for the scope but could be tightened.

Highly actionable throughout: concrete CLI commands for triage queries, specific restore-target syntax, a mitigation table mapping symptoms to exact mutation commands, structured hand-off instructions with `--change-desc` templates, and executable tag/verify commands for close-out. Every path has copy-paste-ready commands and clear routing decisions.

Excellent multi-step workflow with five explicit triage questions that route to three clearly defined paths (A/B/C), each with sequenced steps and validation checkpoints. The 'Stop conditions' section acts as guardrails, the 'Verify chain' section provides explicit post-action validation, and feedback loops are present (e.g., 'confirm the targeted skill returned success before the next step'). The 'After green — close-out' section provides a clear post-incident checklist.

References to companion skills and reference files are well-signaled at the bottom, and the skill appropriately delegates mutations to other skills. However, no bundle files were provided to verify the referenced paths exist, and the skill itself is quite long (~250 lines) with the mitigation table and close-out details that could potentially be split into referenced sub-documents. The content is well-organized with clear headers but is borderline monolithic.