Content
92%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is concise, highly actionable, and sequences a risky workflow with real validation checkpoints. Its only defect is a dangling link to a reference file that is not bundled, which breaks the progressive-disclosure path.
Suggestions
Either create references/reference.md with the promised command snippets and merge message template, or remove the closing link so the skill does not point to a missing file.
If keeping the reference, place it under references/ and update the link path to references/reference.md to match standard bundle layout.
Consider moving the 'Quick reference' ours-vs-theirs staging notes into that reference file to keep the main body even leaner once the file exists.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Lean and assumes Claude's competence — it never explains what a merge or conflict marker is; every section (Goals, Hard rules, Workflow, Anti-patterns, Quick reference) earns its place with actionable content. | 3 / 3 |
Actionability | Concrete, executable commands throughout — 'git fetch', 'git merge origin/main', 'git show :2:path' vs 'git show :3:path', 'git checkout --conflict=merge', 'git merge --abort' — with copy-paste-ready examples rather than pseudocode. | 3 / 3 |
Workflow Clarity | A clear numbered 1–5 sequence with explicit validation checkpoints ('Re-run git status; ensure no conflict markers remain; run tests or lint') and stop-at-conflicts gating, providing the feedback loop the rubric expects for destructive merge operations. | 3 / 3 |
Progressive Disclosure | The body is well-sectioned and signals a one-level-deep reference, but the linked '[reference.md](reference.md)' does not exist — no references/ directory or file is present — so navigation is broken rather than functional. | 2 / 3 |
Total | 11 / 12 Passed |