Conciseness

The body is dense and assumes Claude's competence with no concept-explaining padding, but the router distinction is stated three times (intro 'it is not a fast overview', the dedicated 'How it differs from a quick research router' section, and the Cross-References entry) — trimmable redundancy that keeps it just below the lean/every-token-earns-its-place anchor.

Actionability

Concrete file formats (plan.md, sources/NN_slug.md, refresh_targets.md, sources.csv), explicit thresholds (≥3 independent sources, 2–4 falsifiable hypotheses, 4 self-critique questions, Credibility/Recency/Bias scoring), and a per-phase pipeline give mostly executable guidance; the gaps — the 4 self-critique questions themselves, sub-agent prompts, and a source-file template — are more than trivial but not disqualifying, fitting 'mostly executable guidance with minor gaps' over the incomplete-guidance anchor.

Workflow Clarity

The 9 phases are clearly sequenced in an ordered table with depth gating; validation is explicit (Phase 6.5 'lightweight citation check before closing', Phase 5 'insufficient evidence' flagging, hypothesis confirm/refute); Phase 4's 're-evaluate between rounds' is a feedback loop with error handling ('An empty fetch produces an empty claim, never a fabricated citation'); and the Anti-Patterns section serves as a failure-mode checklist — all top-anchor elements are present.

Progressive Disclosure

The bundle contains references/full-catalog.md holding the entire sourcing surface (29 channels, 39 APIs, 103 blocks), yet the body never mentions or links it — 'references present but not clearly signaled' — and the catalog file itself points to an external GitHub repo, adding a second hop to the actual content. The body's own structure is well organized, so this sits at the some-structure anchor rather than lower.

Specificity

The description lists five concrete, specific actions — 'fan-out web search across many channels, parallel sub-agents, source triangulation (each claim backed by ≥3 independent sources), an adversarial review pass, and every source saved to its own file with verbatim quotes' — with comprehensive coverage and no gaps, matching the anchor for multiple specific concrete actions.

Completeness

It explicitly answers what ('Run a disciplined, multi-source research investigation...') and when ('Use when a low-quality answer is expensive: strategy work, comparing N products/methods/markets, validating a hypothesis with external data, or mapping how a field works') with concrete trigger phrases, plus explicit NOT-for boundaries — clearly the top anchor.

Trigger Term Quality

Natural trigger phrases are well covered ('strategy work', 'comparing N products/methods/markets', 'validating a hypothesis', 'mapping how a field works', 'quick fact-checks', 'topic overviews'), but a few natural synonyms users would say ('due diligence', 'deep dive', 'in-depth investigation', 'research report') are missing, fitting 'good keyword coverage; a few natural terms missing' rather than the comprehensive anchor.

Distinctiveness Conflict Risk

Explicit negative boundary clauses — 'NOT for quick fact-checks (answer directly), structured 12-dimension competitor scoring (use competitive-teardown), or fast topic overviews... (use the research router instead)' — actively route away from sibling skills, giving a clear niche with minimal conflict risk.