Content
81%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 strong operational skill: executable commands, an unambiguous step sequence, and unusually rigorous validation gates around every destructive action (typed confirmation, refused-removal menu, no --force on its own initiative). The only real critique is mild token overhead from the rationalizations/quick-reference tables and a slightly long single-file layout.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body assumes Claude's competence — no git or concept explanations — and stays operational throughout. Minor trim opportunities remain: the 10-row "Common Rationalizations" table and the Quick Reference table repeat points already made inline, keeping it just below "every token earns its place". | 4 / 5 |
Actionability | Concrete, executable bash throughout (git rev-parse environment detection, git worktree remove "$WORKTREE_PATH", branch deletion) with exact menu texts. Small gaps — the "<test command>" placeholder and the deliberately flexible forge-tooling guidance ("its CLI if one is available") — keep it from being copy-paste ready for the common cases. | 4 / 5 |
Workflow Clarity | A clearly sequenced six-step pipeline with explicit validation checkpoints and feedback loops: the test suite gates the menu, the merged result is re-tested with stop-and-investigate on failure, destructive discard requires the typed word "discard", and refused worktree removal triggers a confirm-before-force menu. This is exactly the validate→fix→retry discipline the 5 anchor describes. | 5 / 5 |
Progressive Disclosure | No bundle files exist, and the skill is a well-organized single file with clear sections and no buried or nested references. The ~225-line monolith (e.g., the rationalizations table could live in a reference file) and size above the simple-skill threshold keep it at "good structure, minor organization gaps" rather than 5. | 4 / 5 |
Total | 17 / 20 Passed |