Conciseness

Lean and dense with zero padding: no general concept is explained (IL, boxing, delegates, and JIT behavior are referenced, never taught), and every line carries tool-specific semantics such as "'--top' selects a curated ranked prefix, while '-n'/'--rows' shape output; they are not interchangeable". Every token earns its place.

Actionability

Mostly executable guidance: multiple full 'dnx dotnet-inspect -y -- ...' command blocks with concrete flags and a question→entrance table. Not 5 because a few instructions lack executable form ('Retain Candidate, Finding, provenance and physical Member/IL coordinates' specifies no destination, and 'Load "skill correctness"' gives no command).

Workflow Clarity

A clear progression (start with leverage → drill one exact Member → broader investigations) with real checkpoints: 'Stop when each pursued candidate is confirmed, dismissed or explicitly open' and 'Confirm the adjacent pair with diff --package...'. Not 5 because steps are thematic sections rather than an explicit sequence, and there is no validate→fix→retry feedback loop; not 3 because the sequence and termination criteria are unambiguous.

Progressive Disclosure

Good structure: five well-scoped sections, ~93 lines, self-contained with no bundle files to reference. Not 5 because the >50-line simple-skill exception does not apply, and the pointers that do exist ('Load "skill correctness"', 'Load "skill compatibility"') are external skill mentions rather than clearly signaled one-level-deep references; not 3 since nothing that belongs in a separate file is inlined.

Specificity

Names three concrete, domain-specific actions ('Find high-leverage .NET methods with allocation and repetition evidence', 'trace candidate costs through calls and IL', 'select focused profiling or benchmark follow-ups'). Not 5 because the body's resource/pool triage and history-diff capabilities are absent, leaving minor coverage gaps.

Completeness

The 'what' is explicit and clear (find candidates, trace costs, select follow-ups), but there is no 'Use when...' clause or equivalent explicit trigger guidance — 'rather than claiming measured wins' is a scope caveat, not a when-to-use trigger. Per the rubric guideline, missing trigger guidance caps completeness at 3.

Trigger Term Quality

Good natural keyword coverage: '.NET', 'allocation', 'profiling', 'benchmark' are terms users would actually say. Not 5 because common synonyms like 'C#', 'memory', and 'performance' itself are missing; not 3 since the keyword set goes beyond minimal domain mention.

Distinctiveness Conflict Risk

A clear niche (.NET/IL static analysis for performance candidates) with distinct terms ('.NET methods', 'IL', 'allocation', 'profiling or benchmark') that would not naturally trigger unrelated skills. Anchor 5 fits: minimal conflict risk.