Content
32%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 thin, generic shell: it sketches a rough workflow but provides no concrete prompt-engineering patterns, examples, or evaluation methods, and its single detailed reference points to a nonexistent file. Padding (role assignment, unsourced impact statistics, vacuous trigger sections) occupies space that actionable guidance should fill.
Suggestions
Replace circular instructions like "Apply relevant best practices and validate outcomes" with concrete techniques: e.g., before/after prompt examples, a few-shot pattern template, and a specific comparison method (run both prompt variants on N test cases and compare outputs).
Fix the broken progressive disclosure: either create resources/implementation-playbook.md with the promised "detailed patterns and examples" or inline the essential patterns directly in SKILL.md and remove the dangling reference.
Remove non-instructional padding — the "You are an expert prompt engineer" role sentence, the unsourced 40%/30%/50-80% statistics, and the "Do not use this skill when" bullets that restate the description — and add an explicit validation checkpoint (e.g., how to verify an optimized prompt actually outperforms the original before delivering it).
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is short overall, but contains unnecessary padding: the role-play opener "You are an expert prompt engineer specializing in... constitutional AI, chain-of-thought reasoning", the unsourced marketing statistics ("improve accuracy by 40%, reduce hallucinations by 30%, and cut costs by 50-80%"), and vacuous exclusions like "The task is unrelated to prompt optimization". Mostly efficient with sections that should be trimmed — anchor 3. | 3 / 5 |
Actionability | The guidance is circular — "Apply relevant best practices and validate outcomes" and "Provide actionable steps and verification" restate the task rather than instruct how to do it, with no concrete prompt patterns, examples, or comparison methodology. The only concrete pointer ("open resources/implementation-playbook.md") targets a file that does not exist in the bundle, matching anchor 2's minimal high-level guidance missing specific executable steps. | 2 / 5 |
Workflow Clarity | A rough sequence exists (clarify → apply → provide), but the core step is a placeholder ("Apply relevant best practices" is the entire task restated) and validation is named with no method or checkpoint, fitting anchor 2's rough sequence with poorly defined steps and effectively absent validation. The simple-skill exception does not apply since prompt optimization is inherently multi-step. | 2 / 5 |
Progressive Disclosure | The body defers all detailed content to "resources/implementation-playbook.md", but that file does not exist anywhere in the bundle (no resources/ directory), so the one-level-deep reference is dangling and the disclosure chain is broken at its only link. Body space is also spent on "Use this skill when / Do not use this skill when" sections that duplicate the description's job, matching anchor 2's minimal functional structure. | 2 / 5 |
Total | 9 / 20 Passed |