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 lean diagnostic skill: it captures trigger conditions, verbatim proof, the underlying core semantics, a concrete fix, and a fast verification path in about 15 lines. The only gap is the absence of a copy-paste code snippet for the readEmptyAs fix and the throwaway unit test.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Lean and efficient: five short labeled sections with no padding, and the "Why" section covers obscure Terraform plan-diff semantics (prior-value matching) that Claude would not already know, so every token earns its place. Not below 5 since nothing explained is common knowledge. | 5 / 5 |
Actionability | Concrete guidance throughout — the exact config key "readEmptyAs: \"null\"", verbatim error strings for both top-level and nested cases, the directive "Do not add a plan-side key", and a concrete verification recipe ("a throwaway resource.UnitTest with a one-attribute provider"). It falls short of 5 only because no copy-paste code snippet is provided for the fix or the test. | 4 / 5 |
Workflow Clarity | A single-task skill whose one action is unambiguous (keep [] in the plan; achieve null-on-read via readEmptyAs), with an explicit validation step ("Check: A throwaway resource.UnitTest ... shows it in seconds") — the simple-skill exception applies and no destructive/batch cap is relevant. | 5 / 5 |
Progressive Disclosure | Under 50 lines with well-organized labeled sections (Trigger/Proof/Why/Fix/Check), no bundle files exist, and none are needed — the under-50-lines/no-external-references exception gives this a 5. | 5 / 5 |
Total | 19 / 20 Passed |