Content
35%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 reads as a padded agent system prompt rather than a skill: it inventories expertise areas and persona traits Claude already has instead of teaching anything domain-specific. No worked example, benchmark, or template appears anywhere, and its single reference is a dangling path.
Suggestions
Cut the persona padding (Behavioral Traits, Quality Standards, Integration with Commands, Tools and Resources) and the redundant Knowledge Base section, which repeats Core Expertise.
Fix or remove the dangling reference to `resources/implementation-playbook.md` (no such file exists in the bundle), and move the knowledge-base detail into actual reference files one level deep.
Add at least one worked example with real numbers — e.g., a bottom-up TAM calculation or a CAC/LTV/payback computation with benchmark ranges — so the guidance is executable rather than descriptive.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~316-line body is heavily padded: "Behavioral Traits" ("Data-driven: Always grounds recommendations in data"), "Quality Standards" ("Never make unsupported claims"), and "Knowledge Base" largely repeat "Core Expertise" with generic guidance Claude already knows, and no actual benchmarks, formulas, or templates justify the length. | 2 / 5 |
Actionability | "Clarify goals, constraints, and required inputs. Apply relevant best practices and validate outcomes" is abstract, and Response Approach steps ("Apply frameworks", "Gather necessary data") are high-level hints — CAC, LTV, and burn multiple are named but never demonstrated with a calculation, benchmark range, or worked example. The 10-step Response Approach and Output Format section provide some concrete shape, keeping this above 1. | 2 / 5 |
Workflow Clarity | The Response Approach lists a coherent 10-step sequence including a validation step ("Validate findings - Cross-check with benchmarks and alternatives"), but checkpoints are abstract descriptions rather than pass/fail criteria, and two competing instruction sections (the top "Instructions" bullets vs. "Response Approach") muddy which workflow governs. | 3 / 5 |
Progressive Disclosure | Sections are well-headered, but the only bundle reference — "open `resources/implementation-playbook.md`" — points to a file that does not exist in the bundle, and ~200 lines of knowledge-inventory detail that belongs in separate reference files are inlined; the five "leverages skills" (e.g., market-sizing-analysis) are also absent from the bundle. | 3 / 5 |
Total | 10 / 20 Passed |