Content
65%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-structured, actionable command skill with a clear question framework and output template. It is held back by missing validation checkpoints in the workflow and external, non-bundle references that weaken progressive disclosure.
Suggestions
Add explicit validation checkpoints to the workflow (e.g., 'Confirm contract value/term thresholds from When-to-Run before scoring risk', 'Verify counterparty and jurisdiction before finalizing the verdict') to lift workflow_clarity above 3.
Either co-locate the referenced scanner/analyzer scripts and references inside this skill's ./scripts and ./references with one-level-deep links, or remove the cross-repo ../../../ paths and describe the integration as a pointer rather than an in-bundle dependency.
Trim editorializing lines such as 'a single missed clause costs more than a year of engineering' and fold the not-legal-advice disclaimer into a single concise line to improve token efficiency.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Bullet-driven and largely efficient with minimal padding, though a few editorializing lines ('a single missed clause costs more than a year of engineering') and the disclaimer could be trimmed. | 4 / 5 |
Actionability | Concrete six-question checklist, a complete output-format template with example table rows, a verdict schema, and executable Python tool invocations; minor gaps in showing how tool output feeds the verdict. | 4 / 5 |
Workflow Clarity | A clear four-step sequence (read → run six questions → top-3 issues → verdict) is present, but there are no explicit validation checkpoints or feedback loops (e.g., confirm counterparty scope before assigning risk levels). | 3 / 5 |
Progressive Disclosure | The body is reasonably sectioned, but referenced scripts/references live outside this skill bundle via deeply-relative cross-repo paths (../../../c-level-advisor/...) and no ./references, ./scripts, or ./assets files exist, so references are not clearly signaled as one-level-deep bundle content. | 3 / 5 |
Total | 14 / 20 Passed |