Content
82%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is a well-structured, concise instruction skill with concrete per-mode checklists and a delivery quality gate. It earns high marks on token efficiency and organization; actionability and workflow clarity sit just below top because there are no executable artifacts and no explicit error-recovery feedback loop.
Suggestions
Add an explicit validate→fix→retry feedback loop in the Quality Gate (e.g., 'If a claim lacks a source, re-search before delivering') to lift workflow_clarity toward 5.
Provide one short worked example (e.g., a sample sourced TAM estimate or competitor table) to make the guidance copy-paste ready and raise actionability.
Consider extracting the per-mode collection checklists into a one-level-deep reference file (e.g., MODES.md) signaled from the body to strengthen progressive disclosure.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Lean and efficient throughout: tight bullet checklists, no padding explaining what market research is, and every line earns its place ('Produce research that supports decisions, not research theater.'). It assumes Claude's competence. | 5 / 5 |
Actionability | Provides concrete, specific checklists per research mode ('Collect: fund size, stage, and typical check size'; 'Use: top-down estimates... bottom-up sanity checks'), which is actionable for an instruction-only skill, but lacks copy-paste executable artifacts that would push it to 5. | 4 / 5 |
Workflow Clarity | Clear sequence via numbered Research Standards, a default Output Format (1–6), and a Quality Gate checklist before delivery; the gate is a checkpoint but not a true validate→fix→retry feedback loop, leaving it just below 5. | 4 / 5 |
Progressive Disclosure | Well-organized into clearly labeled sections with no nested or buried references and no bundle files present; content is appropriately self-contained, though all material is inline with no one-level-deep references to split out. | 4 / 5 |
Total | 17 / 20 Passed |