Conciseness

The body is dense with judgment-relevant craft guidance and never tutorials concepts Claude already knows, but the Non-goals section restates gates already stated verbatim in the steps ("Don't assign blame to people" repeats Step 0, "Don't fabricate the timeline" repeats Step 1's gate, "Don't stop at the first plausible cause" repeats Step 6), and a few rhetorical flourishes ("Repeat classes are where postmortems earn their keep") could be trimmed. This matches the level-4 anchor (efficient, minor instances that could be trimmed) rather than 5, because some redundancy and persuasive prose does not earn its tokens; not 3, because there is no padded explanation or tutorial filler.

Actionability

Every step carries copy-paste-ready commands — `exec("ls -A postmortems/")`, `search({ query: "<subsystem> <failure mode>" })`, `write({ document: { path: "postmortems/YYYY-MM-DD-name.md", template: "postmortem" } })`, `audit({ path: "postmortems/YYYY-MM-DD-name.md" })` — plus a complete worked timeline example, a fully-formed action-item line with link syntax, and a runbook-stub write call. This matches the level-5 anchor (fully executable, specific examples covering the common cases).

Workflow Clarity

Steps 0–11 are clearly sequenced, each with concrete sub-guidance, and the skill includes explicit validation checkpoints with a feedback loop (Step 10: "audit(...) returns clean... — fix every finding") plus HARD GATES that act as go/no-go checkpoints at the riskiest moments (blame-seeking framing, unsourced timeline entries). This matches the level-5 anchor (clear sequence with explicit validation steps, feedback loops for error recovery).

Progressive Disclosure

As a single-file skill with no references/, scripts/, or assets/ bundles, the body is cleanly sectioned (Steps 0–11, Non-goals) with no buried or nested references, and all cross-references (to the /open-knowledge skill, guides/ stubs, sibling skills) are clearly signaled one level deep. This matches the level-4 anchor (good structure, content appropriately placed, minor organization gaps) rather than 5, because at ~190 lines — beyond the under-50-line simple-skill exception — nothing is split out and sections like the trigger-vs-root-cause definitions or blamelessness rules could live in a reference file.

Specificity

The description enumerates multiple concrete deliverables — "evidence-based timeline, trigger vs root cause vs symptom, contributing factors, what went well, and owned+dated+verifiable action items" — anchored to a named structure ("the Google SRE shape") with a concrete output location ("under postmortems/"). This matches the level-5 anchor (multiple specific concrete actions, comprehensive coverage); nothing is vague and no capability is over-claimed.

Completeness

It explicitly answers both questions: the first sentence states what the skill does, and "Read when asked to write a postmortem..." explicitly states when to use it with concrete trigger phrases. This matches the level-5 anchor exactly; a 4 would require the 'when' clause to be less explicit, which it is not.

Trigger Term Quality

It covers the natural phrases users would actually say: "write a postmortem", "do an incident review", "run a root cause analysis", "write up the outage", "retro on the outage", and "we had an incident and wants it documented". This matches the level-5 anchor (comprehensive coverage of natural terms including synonyms); it is not below it because no common variation of the request is missing.

Distinctiveness Conflict Risk

It occupies a clear niche (documenting an incident that already happened) and adds explicit negative boundary guidance — "Do NOT read to frame a proposal (use frame-a-proposal), write a spec (use write-a-spec), record a decision (use record-a-decision), or review a design (use review-a-design)" — which routes overlapping requests to sibling skills. This matches the level-5 anchor (clear niche with distinct triggers, minimal conflict risk).