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 lean skill body: every line is an operational directive or executable example, the input/output contract is fully specified and consistent with the bundled script, and structure is appropriately minimal for a single-purpose tool. No material weaknesses found.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~35-line body is entirely operational directives — rules, input schema, output schema, one runnable command — with zero explanation of concepts Claude already knows (e.g., 'Treat displayed `...` in tool previews as UI truncation' is non-obvious, earned guidance). Not a 4 because nothing needs trimming: the repeated search payload appears once as an input pattern and once inside the full executable command, each serving a distinct purpose. | 5 / 5 |
Actionability | Fully executable: a copy-paste-ready `echo ... | python scripts/rest_request.py` command, a complete list of required/optional input fields, two concrete JSON payloads for the common cases (object lookup and search), and a precise output contract ('ok', 'records', 'error.code', 'error.message'). The documented interface matches the actual `scripts/rest_request.py` implementation. Not a 4 because no key execution detail is missing. | 5 / 5 |
Workflow Clarity | This is a simple single-task skill (issue one API request) and the single action is unambiguous: build the JSON input per the documented fields, run the script, interpret success vs. 'ok=false' with 'error.code'/'error.message', and use 'save_raw=true' for full payloads. The failure-output contract plus rules like 'Re-run requests in long conversations' provide the needed checkpoints; the destructive/batch validation cap does not apply since these are read-only lookups. | 5 / 5 |
Progressive Disclosure | Under 50 lines with no need for external reference files, and the body explicitly states 'No additional runtime references are required; keep the import package limited to this file and scripts/rest_request.py' — which accurately matches the actual bundle (SKILL.md plus the single script). Sections (Operating rules, Execution behavior, Input, Output, Execution, References) are well-organized and flat, so the simple-skill exception applies cleanly. | 5 / 5 |
Total | 20 / 20 Passed |