Content
67%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 body is an efficient, well-organized single-file overview with genuinely useful, concrete best-practice anchors (MODEL_INFERENCE_STEPS, resource flags, per-model negative-prompt support). Its main weakness is actionability: the command-location guidance stays at the level of structural hints without any module names, commands, or search heuristics that would let Claude act directly.
Suggestions
Add concrete location hints or a discovery command to 'How to find commands', e.g. actual module/file patterns or a grep strategy like searching for argparse subparsers under the CLI package, instead of only category descriptions.
Include one short example of constructing a CLI call that applies the best practices (checking MODEL_INFERENCE_STEPS defaults, adding --low-ram) so the guidance is demonstrably executable.
Consider moving the per-model capability inventory (model list, negative-prompt support matrix) into a references/ file and linking it from SKILL.md to keep the overview leaner.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is list-driven and lean with no explanations of concepts Claude already knows; only the opening paragraph ("Use this skill to inventory CLI capabilities, summarize what the CLI supports...") slightly repeats the description and a few best-practice elaborations could be trimmed. It matches 'Efficient; minor instances of over-explanation that could be trimmed' rather than score 5's every-token-earns-its-place. | 4 / 5 |
Actionability | Some guidance is concrete ("*always* check `MODEL_INFERENCE_STEPS`", "`--low-ram`", "`--stepwise-image-output-dir`") but the core 'How to find commands' section gives only category descriptions like "Common/shared CLI arguments live in the central CLI parser module" with no module names, commands, or search strategies, leaving key execution details missing. This matches 'Some concrete guidance but incomplete... missing key details' rather than score 4's mostly-executable guidance. | 3 / 5 |
Workflow Clarity | For a lookup/navigation skill the sections are unambiguous: 'When to use' enumerates trigger cases, the structure map tells where each capability class lives, and the best-practices rules order command construction (check step defaults, mention resource flags, update READMEs in the same pass). It matches 'Clear sequence with most checkpoints present; minor validation gaps' — the 'how to find' guidance is a heuristic list rather than an ordered procedure, and no verification step exists for confirming a located command still exists. No destructive/batch cap applies. | 4 / 5 |
Progressive Disclosure | There are no bundle files and no nested references; the single file is well-sectioned (When to use, How to find, Best practices, Capability inventory) and everything is one level deep. It matches 'Good structure; most content is appropriately placed; references mostly clear; minor organization gaps' — the body is ~80 lines so the under-50-line simple-skill exception does not apply, and the per-model capability inventory (e.g. the negative-prompt support matrix) could arguably live in a reference file. | 4 / 5 |
Total | 15 / 20 Passed |