Content
92%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 lean, fully sequenced six-step workflow with exact npm commands, an explicit failure-classification scheme, and a defined output contract. Minor gaps in executable detail ('relevant logs', 'active repo-owned processes') keep actionability just short of perfect, but token efficiency and structure are excellent for a skill of this size.
Suggestions
Name the specific log files or locations to inspect in step 1 and how to identify 'active repo-owned processes' (e.g. a ps/pgrep pattern) to close the actionability gap.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Six lean steps with no concept explanations or padding; every line instructs and assumes Claude's competence ('Inspect branch, status, worktrees, runtime versions...'). | 5 / 5 |
Actionability | Exact commands ('npm run check:runtime', 'npm run check:skills', 'npm run verify:cheap'), an enumerated six-way failure taxonomy, and a defined output format make most steps executable; minor gaps remain in 'relevant logs' and 'active repo-owned processes', which specify neither locations nor identification method. | 4 / 5 |
Workflow Clarity | A clear 1-6 sequence in which the named checks are explicit checkpoints, step 4 is the failure-handling path (classify into product code, test, tooling, resources, environment, or provider-gated), and step 6 defines the output contract with a guardrail in step 5. | 5 / 5 |
Progressive Disclosure | At 13 lines with no external references needed, the flat numbered list under a single header is appropriately organized; per the simple-skill guidance, under-50-line skills with no reference needs can score 5 with well-organized sections. | 5 / 5 |
Total | 19 / 20 Passed |