Conciseness

Lean and free of basic-concept padding, assuming Claude's competence; the Step 5 citation and quote block runs somewhat long, which keeps it just below a 5.

Actionability

Provides a concrete SurvivedMutant interface, per-mutator code examples, and a copy-paste-ready proposed-test block, plus a four-part completeness checklist for proposals.

Workflow Clarity

Steps 1–5 are clearly sequenced with explicit checkpoints ('did any covering test already run the separating input?') and verification guidance ('confirm the intended cap semantics against the specification before changing either side'); the work is read-only so the destructive-cap does not apply.

Progressive Disclosure

Overview stays in SKILL.md while per-tool field mappings and operator-name tables are moved to a single, well-signaled, one-level-deep reference (references/tool-normalization.md) that exists in the bundle.

Specificity

Names multiple concrete actions — 'Normalizes a surviving-mutant record', 'classifies why it survived', 'applies per-mutator heuristics', 'drafts the specific test that would kill it' — covering the workflow comprehensively.

Completeness

Explicitly answers both what (normalize/classify/heuristics/draft) and when — 'Use when a mutation run has finished and the report lists surviving mutants (typically 5+)' is a concrete trigger phrase.

Trigger Term Quality

Good coverage of natural domain terms ('surviving mutants', 'mutation run', 'mutation report', all four tool names), though a few common phrasings like 'mutant not killed' or 'mutation score' are absent.

Distinctiveness Conflict Risk

A very specific niche (surviving-mutant triage across named mutation tools) with distinct triggers and minimal overlap with other testing skills.