Content
88%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 unusually disciplined workflow skill: fully executable commands, a clearly sequenced process with genuine validation and fallback steps, and no filler. It is slightly over-dense in places and keeps all policy inline rather than splitting peripheral detail into reference files.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense, imperative policy Claude would not know ("Stage named paths, never `git add -A`") with no conceptual padding, but sections like Screenshots run long and a few justifications (e.g., the three-dots diff explanation) could be trimmed — matching "efficient; minor instances of over-explanation" rather than "every token earns its place". | 4 / 5 |
Actionability | Guidance is copy-paste ready throughout: the batched context-gathering block, `gh pr create --base <base> --title "<title>" --body-file <path> --assignee @me`, the executable base-selection count via `git rev-list --count $(git merge-base ...)`, a real script (`"$SKILL_DIR/scripts/pr-stack.sh"`), and a concrete Too long/Right example table. | 5 / 5 |
Workflow Clarity | A clear sequence (Gather context → Confirm before acting → Create or update → Description format → Cut) with explicit validation checkpoints and error-recovery loops: "Wait for approval", "Check `git status` after staging; if anything unexpected appears, stop and ask", "view each image before attaching; one that looks wrong is a bug to fix", "If an upload fails, the PR still opens — tell the user which are missing", and the `gh stack --help` failure fallback. | 5 / 5 |
Progressive Disclosure | Sections are well-organized and the single bundle file (scripts/pr-stack.sh, verified to exist) is referenced one level deep and clearly signaled via `SKILL_DIR`; however, ~125 lines of policy are fully inlined where self-contained detail (the Screenshots rules, the base-selection algorithm) could live in reference files, matching "good structure; minor organization gaps" rather than the ideal split. | 4 / 5 |
Total | 18 / 20 Passed |