Content
25%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.
The body is well-sectioned but nearly content-free: generic instruction boilerplate fills the space where a concrete PR-enhancement workflow should be, and the sole reference to detailed material points at a nonexistent file. Claude would finish reading it knowing the desired output categories but not how to accomplish any of them.
Suggestions
Replace the generic Instructions with a concrete PR-enhancement workflow, e.g. inspect the diff (git diff main...HEAD or gh pr diff), draft the description from a stated template, run and compare test coverage, and propose split points for oversized PRs.
Include a filled-in example PR description or a copy-paste template under Output Format so the eight deliverable categories become executable guidance.
Fix or remove the dangling resources/implementation-playbook.md reference — either ship the playbook with worked patterns or delete the pointer so Claude is not directed to a missing file.
Cut the description-duplicating intro and the tautological Use/Do-not-use/Limitations boilerplate; state one domain-specific trigger line and one real limitation instead.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Noticeably padded: the opening paragraph duplicates the frontmatter description verbatim, and "Use this skill when / Do not use this skill when / Limitations" sections are tautological boilerplate ("The task is unrelated to pull request enhancement") that teaches Claude nothing PR-specific. This matches anchor 2 (several unnecessary or padded sections) better than anchor 3, since well over half the body is generic scaffolding. | 2 / 5 |
Actionability | The Instructions are entirely generic — "Clarify goals, constraints, and required inputs. Apply relevant best practices and validate outcomes" — with no commands (e.g. gh pr), no PR description template, and no example. The Output Format section enumerates eight deliverable names (PR Summary, Review Checklist, Risk Assessment, etc.), which is the only concrete element, but it stays at the level of high-level hints about what to produce with zero guidance on how, matching anchor 2. | 2 / 5 |
Workflow Clarity | There is only a rough pseudo-sequence (clarify goals → apply best practices → provide steps and verification) with no domain-specific steps (analyze the diff, run coverage, draft the description, propose splits) and no validation checkpoints, matching anchor 2 (rough sequence, steps poorly defined, validation absent) rather than anchor 3's more defined step list. | 2 / 5 |
Progressive Disclosure | The body is short with clear section headers, but its only depth pointer — "open resources/implementation-playbook.md" and the matching Resources entry — references a file that does not exist in the bundle (no resources/, references/, scripts/, or assets/ directories). The promised "detailed patterns and examples" are therefore unreachable, so navigation to the substantive content fails outright, placing this below anchor 3. | 2 / 5 |
Total | 8 / 20 Passed |