Content
93%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 exemplary lean skill: fully executable commands, a clear guarded workflow with pre-flight and post-hoc verification, and no wasted tokens. The single gap is the absence of a rollback/recovery path for a destructive history-rewriting operation.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Lean and command-driven throughout: no explanation of git concepts Claude already knows, and every line ("Stop if the working tree is dirty", "never use `git rebase -i`") earns its place. | 5 / 5 |
Actionability | Seven concrete, executable commands (git rev-parse, git status --porcelain=v1, git show, git reset HEAD~1, git add, git commit, git log) cover the entire flow; the only placeholders (<files>, commit message template) are inherent user-specific parameters, not pseudocode. | 5 / 5 |
Workflow Clarity | Eight clearly sequenced steps with explicit validation checkpoints (dirty-tree stop, scope guard, confirmation before rewriting history, post-verification), but no error-recovery or rollback guidance (e.g., restoring via ORIG_HEAD) for this history-rewriting operation, which the score-5 anchor expects. | 4 / 5 |
Progressive Disclosure | A small single-purpose skill (~55 lines) with no bundle files and well-organized sections (Workflow, Constraints); per the rubric's simple-skill exception, well-organized sections alone merit a 5. | 5 / 5 |
Total | 19 / 20 Passed |