Content
86%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 tight, actionable skill body with concrete CLI commands, explicit safety boundaries, and clean section structure. The only real weakness is a slightly over-detailed architecture preamble and the absence of a retry loop in the workflow.
Suggestions
Trim the opening paragraph's internal-implementation detail (SQLite projection, snapshot authority, worker limits) to the constraints Claude actually needs to honor.
Add an explicit fix-and-retry step for non-zero exit codes (e.g., 'on failure, surface the bounded error and re-run with corrected arguments; never fall back to filesystem reads') to close the workflow feedback loop.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly lean and free of generic concept explanation, but the opening paragraph over-shares internal architecture ("process-private SQLite projection", "ordered snapshot authority", worker hard-limits) that Claude does not need to invoke the commands. | 4 / 5 |
Actionability | Fully executable, copy-paste-ready commands with concrete argument syntax (e.g., `minutes people merge <canonical> <alias...>`, `minutes person "<name>"`) covering the common cases. | 5 / 5 |
Workflow Clarity | A clear 4-step sequence with real validation checkpoints (require exit status 0, confirm before merging, hard-stop on error), but it lacks an explicit fix-and-retry feedback loop for failed commands. | 4 / 5 |
Progressive Disclosure | Under 50 lines with no need for external references and well-organized into clearly signaled sections (Privacy boundary, Available commands, Workflow, Output, Gotchas), matching the simple-skill exception for a top score. | 5 / 5 |
Total | 18 / 20 Passed |