Content
100%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.
An exemplar content body: terse, immediately executable, with safety rails surfaced for the destructive operation and a clean single-level reference to the bundled CLI. No padding, no nested references, no over-explanation.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | A lean ~37-line body that leads with executable commands; the prose in Conventions/Gotchas conveys only non-obvious behavior (auto-exclusion from git status, branch/dir-name flattening, default base, main-checkout operation) and never explains concepts Claude already knows. | 5 / 5 |
Actionability | All three subcommands are given as copy-paste-ready `uv run ${CLAUDE_PLUGIN_ROOT}/...` invocations with their flags, and the Resources section points to `--help` for the exact remainder — fully executable guidance covering the common cases. | 5 / 5 |
Workflow Clarity | Each operation is a single unambiguous command; the destructive `remove` path has explicit built-in validation (refuses if dirty, --force discards) with recovery guidance ("Never force without first investigating what would be lost"), satisfying the destructive-operation validation requirement. | 5 / 5 |
Progressive Disclosure | Under 50 lines, well-sectioned (Run it / Conventions / Gotchas / Resources), with a single one-level-deep reference to scripts/worktree.py (verified to exist with matching flags) clearly signaled as "run it, don't read it" — the overview/detail split is appropriate. | 5 / 5 |
Total | 20 / 20 Passed |