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 efficient, actionable body for a simple single-purpose skill: concrete commands, an explicit hard-fail contract, and clean deferral to a source-of-truth doc. Only minor tightening is possible in the closing paragraph's repetition of the trigger condition.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~38-line body is lean: it defers rationale to docs/schemathesis.md ("Não copie essa página") and gives commands without padding. Not 5: the closing paragraph restates the trigger condition already in the frontmatter, and "Uma tarefa que muda a API não está pronta só com o loop de lint" could be trimmed to one clause. | 4 / 5 |
Actionability | Two copy-paste-ready invocations are given ("apm run schemathesis --param spec=<...>" and "python3 -m schemathesis run <caminho-ou-url-do-openapi>") with a clear directive on what to target and an anti-shortcut guard ("Não invente um spec na hora"). Commands cover the common cases for this single-purpose skill. | 5 / 5 |
Workflow Clarity | For a simple single-purpose skill the action is unambiguous: run Schemathesis against the platform's spec before closing the task, with the outcome defined ("Falha dura se a implementação divergir do contrato") and failure guidance (never fabricate a spec to mask divergence). The run itself is the validation checkpoint. | 5 / 5 |
Progressive Disclosure | Short, well-sectioned body under 50 lines with no bundle files needed; the source-of-truth page (docs/schemathesis.md) and complementary layer docs are clearly signaled one level deep at the top. Simple-skill structure is fully appropriate. | 5 / 5 |
Total | 19 / 20 Passed |