Content
100%Weight 40%Scale 1-3Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
An efficient, action-first body that wraps a bundled CLI with executable commands, explicit safety rails on the destructive operation, and clean delegation of detail to the script's own --help. It is well within token budget and leaves nothing essential implicit.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~37-line body is lean and assumes Claude knows git/worktree basics (it never explains what a worktree is), with each section — Run it, Conventions, Gotchas, Resources — earning its place; the only rhetorical flourish ("instead of improvised git worktree incantations") is purposeful, not padding. | 3 / 3 |
Actionability | Provides fully executable, copy-paste-ready commands — "uv run ${CLAUDE_PLUGIN_ROOT}/skills/prp-worktree/scripts/worktree.py create <name> [--base <branch>]" — with concrete flags and per-subcommand behavior, matching the executable-and-specific anchor rather than pseudocode. | 3 / 3 |
Workflow Clarity | Per the simple-skills note the single action (run the CLI) is unambiguous, and the destructive remove path has an explicit validation gate — the script "refuses if the worktree has uncommitted changes" plus "Never force without first investigating what would be lost" — so validation is present, not missing, avoiding the cap at 2. | 3 / 3 |
Progressive Disclosure | Under 50 lines with well-organized sections and a single one-level-deep reference to scripts/worktree.py (verified present), delegating flag detail to "--help on any subcommand" rather than duplicating it inline, matching the clear-overview / well-signaled-reference anchor. | 3 / 3 |
Total | 12 / 12 Passed |