Content
100%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 instruction-only skill body: tight, parameter-specific tool guidance with explicit edge-case and data-integrity handling (truncation, empty results, dataset lag, fatality caveats). Nothing is padded, and every step is unambiguous and executable.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and dense with operational specifics (e.g., "since_days: 14", "limit: 0, so event counts are not cut at the default 30 per list") and contains no explanations of concepts Claude already knows; every line carries operational or data-integrity value. | 5 / 5 |
Actionability | Tool calls are fully specified with concrete parameters and values ("country` with its English name (for example \"Sudan\")", "alerts_only: true", ISO alpha-2 codes, lat/lon/radius_km), plus conditional flags like min_fatalities, so guidance is directly executable for an instruction-only skill. | 5 / 5 |
Workflow Clarity | A clear 4-step sequence with explicit feedback loops: retry without alerts_only when no alert stories return, disclosure when output is truncated or simplified, and an empty-result interpretation rule; the read-only nature means the destructive/batch cap does not apply. | 5 / 5 |
Progressive Disclosure | The skill is under 50 lines, self-contained with no external references needed, and well organized into Tools, Steps, and Rules sections, satisfying the simple-skill exception for progressive disclosure. | 5 / 5 |
Total | 20 / 20 Passed |