Content
67%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 clearly written, well-structured advisory skill with concrete hook setup, configuration, and decision tables. Its main weakness is that the core script it instructs Claude to wire up (suggest-compact.js) is not present in the bundle, undercutting actionability.
Suggestions
Ship the referenced suggest-compact.js in the scripts/ bundle so the hook setup is fully executable end-to-end.
Tighten the Configuration section by removing the blockquote that restates the auto-detection logic already covered in 'How It Works'.
Add one or two concrete example compact invocations (e.g. '/compact Focus on implementing auth middleware next') earlier in the body to anchor the Best Practices guidance.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly lean with task-relevant detail (thresholds, env vars, decision tables) and little generic explanation; a few spots like the Configuration blockquote restate prior points and could be trimmed. | 4 / 5 |
Actionability | Provides copy-paste hook JSON and concrete env-var defaults, but the central artifact it names ('suggest-compact.js') is referenced in the body yet absent from the empty scripts/ bundle, leaving a real gap in executable completeness. | 3 / 5 |
Workflow Clarity | Decision guidance is clearly sequenced via the phase-transition and what-survives tables, giving an unambiguous when-to-compact logic; this advisory skill has no destructive/batch step requiring validation, and checkpoints are implicit but adequate. | 4 / 5 |
Progressive Disclosure | Well-organized with clear section headers and a self-contained overview plus an external 'Related' section; the reference to a missing suggest-compact.js script is a minor structural gap that keeps it below 5. | 4 / 5 |
Total | 15 / 20 Passed |