Content
85%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is an excellent lean procedural guide: tightly sequenced phases, a grounding feedback loop, concrete output structure, and no token waste on concepts Claude already knows. The main gap is actionability — as an instruction-only skill it is specific but stops short of executable artifacts or fully spelled-out tool invocations.
Suggestions
Add one or two concrete tool-invocation examples (e.g., a sample Grep pattern or a mermaid snippet) to make the workflow more copy-paste ready.
Show a brief example of a finished gap/question output entry so the expected format is unambiguous rather than only described.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and assumes Claude's competence — it gives procedural guidance without explaining what a spec, OAuth, or mermaid diagram is, and every section earns its place with concrete directives. This matches the score-3 anchor of lean, efficient content that assumes Claude's intelligence. | 3 / 3 |
Actionability | The guidance is concrete and well-structured (named checklist items, a Good/Bad question example, severity tiers) but it is instructional rather than executable — there is no code, command, or copy-paste artifact, and several steps ('use the native content-search tool', 'walk through the spec as a user') leave the exact execution implicit. This matches the score-2 anchor of concrete guidance that is incomplete and lacks executable specifics. | 2 / 3 |
Workflow Clarity | The four phases are explicitly sequenced (Ground -> Map -> Find -> Formulate) with a defined Output Format, and Phase 1 establishes a validation checkpoint ('Gaps are only gaps if the codebase doesn't already handle them') that feeds back into Phase 3. The multi-step process is clearly ordered with a feedback loop, matching the score-3 anchor. | 3 / 3 |
Progressive Disclosure | The skill is self-contained with no bundle files and the content is well-organized into clearly headed sections (Phases, Output Format, Principles) with no nested references. Per the rubric's simple-skills note, well-organized self-contained content scores 3 even without external file references. | 3 / 3 |
Total | 11 / 12 Passed |