Content
86%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.
The skill body is a tight, well-organized reference: concise mode table, concrete JSON examples, explicit anti-patterns, and clear boundaries against sibling skills. It loses only marginal points for lacking a returned-shape example and an explicit validation checkpoint on schema mode.
Suggestions
Add one short example showing the JSON object returned by `mode: "structured"` so users know what shape to expect.
Make the schema-mode verification (hallucinated values → null) an explicit step the caller can check, e.g. 'verify returned values; nulls signal unmatched fields'.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean: no re-explanation of what HTML/JSON-LD/microdata are, just terse mode descriptions and copy-paste JSON snippets, so every token earns its place. | 5 / 5 |
Actionability | Quick Reference gives concrete, copy-paste-ready JSON invocations for every mode and the named_schema list is explicit, but several modes lack a runnable example showing the returned shape, leaving a minor gap. | 4 / 5 |
Workflow Clarity | Selection guidance is crisp — a mode table, an explicit 'always use structured' rule, an Anti-Patterns list, and a 'When NOT to use' section sequence decisions well, though there are no explicit validate/retry checkpoints for the schema-mode verification step. | 4 / 5 |
Progressive Disclosure | Short, well-sectioned overview with one-level-deep pointers to sibling skills in See Also; no bundle files exist to mis-navigate, and structure is clear and easy to scan. | 5 / 5 |
Total | 18 / 20 Passed |