Content
81%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 lean, highly actionable design guide with a complete lifecycle workflow, explicit verification checkpoints, and debugging recovery paths. Its main structural limitation is that all depth lives inline in SKILL.md with only unverifiable cross-skill references, which is acceptable at this size but leaves no room to grow.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Dense, opinionated guidance with one-line examples ("take a product ID and quantity and calculate the price") that assumes competence — no re-explanation of JSON Schema, WP_Error, or hooks. A few sentences could be tightened ("The information available to each caller helps you choose the inputs the ability should support"), so it sits at efficient-with-minor-trimming rather than every-token-earns-its-place. | 4 / 5 |
Actionability | Concrete, executable patterns throughout: `additionalProperties: false`, a `selector` object using `oneOf`, `content.raw`/`content.rendered`, `{object}-{action}` naming, `meta.annotations` values, and a fully specified `errors` array shape. Per the instruction-only scoring note absent code is not penalized, but a few rules remain abstract ("Choose a stable result type and structure"), keeping it below fully-executable. | 4 / 5 |
Workflow Clarity | The procedure flows from caller analysis through input design, result shape, errors, CRUD consistency, annotations, and permissions, backed by an explicit Verification section (zero/one/many matches, invalid input, denied access, partial-update field checks) and a Failure modes section with symptom-to-diagnosis recovery paths. This matches the anchor with explicit validation steps and error-recovery feedback loops. | 5 / 5 |
Progressive Disclosure | Well-organized sections with clearly signaled, purpose-stated outbound references ("For registration, exposure, and client usage, see [wp-abilities-api]..."), but those references point outside the bundle and could not be verified to resolve, and the ~100-line body exceeds the under-50-line simple-skill exception for scoring structure alone. | 4 / 5 |
Total | 17 / 20 Passed |