Conciseness

The 12-line body is lean and assumes Claude's competence — it never explains what a primary source is beyond a compact parenthetical, what Markdown is, or how background agents work. Every line either directs action or constrains it, matching the 'every token earns its place' anchor.

Actionability

The guidance is mostly executable: 'Spin up a background agent', the enumerated criteria for primary sources ('official docs, source code, specs, first-party APIs'), 'Write the findings to a single Markdown file, citing each claim's source', and a concrete save-location rule. It is not a 5 because it stops short of copy-paste-ready specifics — no example agent prompt, no suggested file-naming pattern or findings-file structure — leaving minor gaps, and not a 3 because the steps are concrete directives rather than high-level hints.

Workflow Clarity

The three numbered steps form a clear, correctly-ordered sequence (investigate → write findings → save to conventional location) with explicit decision guidance ('if there is none, put it somewhere sensible and say where'). It is not a 5 because there is no validation checkpoint (e.g., confirm the agent's file exists or that claims actually cite sources), and not a 3 because the sequence is unambiguous with only that minor gap — no destructive or batch operation is involved, so no cap applies.

Progressive Disclosure

The skill is under 50 lines with no need for external references — no references/, scripts/, or assets/ exist in the bundle, and nothing in the body points to missing files — and the body is well organized as a directive followed by a three-item job list. Per the rubric's simple-skill guideline, this matches the top anchor for a self-contained skill.

Specificity

The description names the domain and two concrete actions — "Investigate a question against high-trust primary sources" and "capture the findings as a Markdown file in the repo" — which matches the anchor for naming the domain with 1-2 concrete actions. It is not a 4 because it stops short of listing several specific actions (e.g., verifying citations, following claims back to owners, organizing output), and not a 2 because the actions it does state are concrete rather than generic.

Completeness

It explicitly answers what ("Investigate a question against high-trust primary sources and capture the findings as a Markdown file in the repo") and when ("Use when the user wants a topic researched, docs or API facts gathered, or reading legwork delegated to a background agent") with concrete trigger phrases, matching the top anchor exactly. It is not a 4 because the 'when' clause is already explicit and specific, listing three distinct trigger scenarios.

Trigger Term Quality

"Use when the user wants a topic researched, docs or API facts gathered, or reading legwork delegated to a background agent" covers good natural phrases — "topic researched", "docs or API facts gathered", "reading legwork" — that users would plausibly say. It is not a 5 because common variations like "look this up", "find out about", "dig into", or specific artifact terms are missing, and not a 3 because multiple natural trigger phrasings are present rather than a few keywords with missing synonyms.

Distinctiveness Conflict Risk

"Investigate a question against high-trust primary sources", "capture the findings as a Markdown file in the repo", and "delegated to a background agent" carve out a recognizable niche of delegated repo-documented research with minimal overlap risk. It is not a 5 because 'research' and 'docs' as triggers could overlap with general web-research or documentation-writing skills, and not a 3 because the primary-sources constraint and file-output requirement make it largely distinct.