Content
63%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 well-structured and actionable with concrete commands and a clear sequenced workflow, but it suffers from restated rules across multiple sections (conciseness) and — more seriously — references 12 bundle files that are absent from the skill, breaking progressive disclosure and the runnability of the tool commands.
Suggestions
Ship the referenced bundle files (references/operating_prompt.md, the canon files, scripts/*.py, assets/*.md) so the progressive-disclosure structure actually resolves and the CLI commands run.
Collapse the Hard Rules and Anti-Patterns sections into the Operating Prompt / Workflow where the rules are already stated, eliminating the repeated counterargument-first, confidence-level, and no-sycophancy rules.
Verify the script argument flags documented in the examples (e.g. --ellis-pct, --must-do) match the actual scripts' interfaces once they are added.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly efficient persona-specific content Claude would not already know, but the same rules are restated across three sections — "Lead with the strongest counterargument" appears in Operating Prompt, Workflow step 4, and Hard Rules #3, and the no-sycophancy / confidence-level rules are repeated verbatim in Operating Prompt, Hard Rules, and Anti-Patterns — which is padding that could be tightened. | 3 / 5 |
Actionability | Provides concrete executable CLI commands with real arguments (e.g. "python scripts/market_first_evaluator.py --size 8 --growth 7 --timing 9 --pull 8 --team 6 --product 5") and a numbered interrogation with recommended answers; the gap is that the referenced scripts/ files are not present in the bundle, so the commands are not actually runnable as-is. | 4 / 5 |
Workflow Clarity | A clearly sequenced four-step workflow (route → interrogation one-at-a-time → score with tools → verdict in voice) with deterministic checkpoints (sub-4 market hard kill gate, Sean Ellis 40% gate, 3-5 cap); minor validation gaps keep it below a 5. | 4 / 5 |
Progressive Disclosure | The body describes a well-signaled, one-level-deep structure across references/, scripts/, and assets/ with one-line descriptions for each file, but none of those 12 referenced files actually exist in the bundle — so navigation leads to missing files and the disclosure structure does not resolve in practice. | 3 / 5 |
Total | 14 / 20 Passed |