Content
92%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 skill body: a checkpoint-driven workflow, terse expert-level operational rules, and clean branch-based navigation to three real one-level-deep reference files, plus a runnable focused-test command. The only minor gap is that most common-case procedures (mock setup, capture helpers) live in the referenced files rather than the body itself.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and assumes competence — e.g. 'Install the mock before navigation. Act through visible controls and assert both the rendered outcome and relevant Gateway requests' — with zero concept explanations or padding; every section delivers operational rules. Not 4 because there is no over-explanation to trim: dense repo-specific facts (fixture helper path, artifact function, suite owners) are stated once, tersely. | 5 / 5 |
Actionability | Provides concrete, executable guidance: a copy-paste-ready command ('node scripts/run-vitest.mjs run --config test/vitest/vitest.ui-e2e.config.ts ...'), a named fixture helper ('ui/src/test-helpers/control-ui-e2e.ts'), and an exact function signature ('createControlUiE2eArtifactDir(scope, parentDir?)'). Not 5 because the most common operations — mock installation, stress-test gallery build, screenshot/trace capture — are delegated to reference files rather than shown, so the body alone doesn't cover the common execution cases. | 4 / 5 |
Workflow Clarity | The four-step workflow is clearly sequenced with an explicit validation checkpoint on every step ('Done: every changed behavior has a named scenario and proof boundary', 'Done: each scenario reaches its asserted outcome, or has a retained failure identifying the unmet assertion'), plus an error-recovery rule ('Resolve unexpected behavior at the product or fixture owner before replaying'). Not 4 because checkpoints are explicit, not implicit, and the failure/retention path is a concrete feedback loop. | 5 / 5 |
Progressive Disclosure | The body is a concise overview with well-signaled, one-level-deep references — [visual proof](references/visual-proof.md), [interaction profiling](references/interaction-performance.md), and [Run Inspector capture](references/run-inspector.md) — all of which exist as real files in the bundle, including a valid anchor ('references/visual-proof.md#ui-stress-test' matches the '## UI Stress Test' heading). Not 4 because navigation is unambiguous per outcome type and no content that belongs in references is inlined. | 5 / 5 |
Total | 19 / 20 Passed |