Content
100%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.
A well-engineered instruction skill: lean, actionable, with a clear sequenced workflow, explicit validation checkpoints, and clean one-level-deep progressive disclosure backed by real bundle files.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is instruction-dense with no padding about concepts Claude already knows (no 'what is CI' or 'what is AGENTS.md' explanations); every section — Protected Invariants, Reference Routing, Workflow — earns its tokens. | 5 / 5 |
Actionability | Provides executable script invocations ('scripts/inspect_project_guidance.py --cwd <repo>'), a copy-paste ready markdown output contract with a changes table, and a concrete promotion-matrix mapping of lesson types to destinations. | 5 / 5 |
Workflow Clarity | A clearly sequenced six-step workflow (Capture → Inspect → Choose Layer → Account for CI Cost → Report First → Apply) with validation checkpoints: 'Re-read target files immediately before editing', 'Keep patches narrow and reversible', and 'Run the smallest relevant validation and report what passed'. | 5 / 5 |
Progressive Disclosure | SKILL.md is a well-signaled overview: the Reference Routing section names each one-level-deep reference with its use case, and every referenced file (references/*.md, scripts/*.py) exists in the bundle, making navigation easy. | 5 / 5 |
Total | 20 / 20 Passed |