Content
93%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, well-structured skill body: executable commands, a clear cheap-to-expensive search discipline, and a diagnostic recovery path. The only mild gap is that workflow validation checkpoints are implicit rather than explicit.
Suggestions
Make the confirm-hits checkpoint explicit in the Progressive search method (e.g. 'Only proceed to the narrow pass once step 1 surfaces a relevant title') to push workflow clarity toward 5.
No changes needed for conciseness, actionability, or progressive disclosure.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and assumes Claude's competence — "Search the user's persistent Claude-Mem memory on cmem.ai" plus a brief, domain-specific note on what memory stores; no generic concept explanation or padding, matching the level-5 anchor. | 5 / 5 |
Actionability | Provides copy-paste-ready commands covering the common cases — `node "${CLAUDE_PLUGIN_ROOT}/scripts/cmem-hook.mjs" search "your query" --limit 20` and the `status` diagnostic — fully executable, matching the level-5 anchor. | 5 / 5 |
Workflow Clarity | The 3-step Index → Narrow → Answer sequence is clear and the Diagnostics section gives an error-recovery path, but validation checkpoints are implicit (skim/confirm hits) rather than explicit validate-then-proceed steps, so it sits at level 4 rather than 5; the operation is read-only so the destructive-cap does not apply. | 4 / 5 |
Progressive Disclosure | A self-contained skill under 50 lines with no in-bundle reference files, organized into well-signaled sections (How to search, Progressive search method, Diagnostics, Notes); the referenced cmem-hook.mjs is a plugin runtime script outside this bundle, so per the simple-skill note this earns a 5. | 5 / 5 |
Total | 19 / 20 Passed |