Content
81%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.
A well-crafted orchestration skill: crisp five-step process, up-front validation of the fixed point, verbatim sub-agent briefs, and disciplined aggregation rules. Remaining gaps are minor — slightly redundant smell definitions, a dangling reference to an external "tracker doc" workflow, and an inlined smell baseline that could be a reference file.
Suggestions
Condense each smell entry to name + fix (drop the "what it is" half Claude already knows), e.g. "Feature Envy → move the method onto the data it envies", and trim "Why two axes" to the two pass/fail bullets.
Replace "fetched via the workflow in the tracker doc" with the actual fetch command or a bundled reference path so the spec-source step is executable from this skill alone.
Move the 12-smell baseline to `references/smells.md` and have step 4 read it and paste it into the Standards sub-agent prompt, keeping SKILL.md as a lean overview.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean — exact git commands, one-line smell entries, verbatim sub-agent briefs, and no padding about git or code review generally. Not 5: each smell entry restates "what it is" ("a method that reaches into another object's data more than its own") that Claude already knows from Fowler's catalogue, and the 8-line "Why two axes" rationale could be trimmed to two lines — minor over-explanation that could be tightened. | 4 / 5 |
Actionability | Concrete and executable throughout: `git diff <fixed-point>...HEAD`, `git log <fixed-point>..HEAD --oneline`, `git rev-parse <fixed-point>`, a numbered spec-source resolution order, and two quoted copy-paste-ready sub-agent briefs with word limits. Not 5: "fetched via the workflow in the tracker doc" points at a document that is neither part of this skill nor bundled, leaving the spec-fetch step non-executable from this file alone — a minor gap. | 4 / 5 |
Workflow Clarity | Five clearly sequenced steps with an explicit validation checkpoint up front ("confirm the fixed point resolves (`git rev-parse <fixed-point>`) and the diff is non-empty. A bad ref or empty diff should fail here, not inside two parallel sub-agents") and error-recovery branches (ask for the fixed point if unspecified; ask where the spec is; Spec sub-agent skips and reports "no spec available"). Matches the anchor-5 pattern of explicit validation plus feedback loops. | 5 / 5 |
Progressive Disclosure | No bundle files exist, and the single-file body is well organized into clearly headed sections (Process, Why two axes) with each step's content in its right place. Not 5: the ~15-line smell baseline is inlined reference data that could live in a `references/` file the orchestrator reads and pastes into the sub-agent prompt; the skill is also over the ~50-line simple-skill zone, so the inlining is a minor organization gap rather than ideal. | 4 / 5 |
Total | 17 / 20 Passed |