Content
50%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 has a solid core — accurate, executable CLI examples for all three capabilities — wrapped in a large shell of generic template boilerplate. The main problems are padding, duplicated sections, inconsistent script paths, and a Quick Validation section whose expected output contradicts the actual script behavior.
Suggestions
Delete the generic template sections (When to Use, Key Features, Dependencies, Implementation Details, Required Inputs, Recommended Workflow, Output Contract, Validation and Safety Rules, Failure Handling, Deterministic Output Rules, Completion Checklist) — they restate what Claude already knows and more than double the token cost of the real content.
Unify script paths to `scripts/client.py` throughout and remove the phantom `cd "20260316/scientific-skills/…"` and `.trae/skills/…` paths so the Usage examples are copy-paste runnable.
Fix the Quick Validation section to describe what `python scripts/client.py --help` actually prints (argparse usage), and document the supported `--format` (json/xml/txt/tsv) and `--sort` options.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Over half the body is generic template filler that assumes no intelligence ("Use this skill when the request matches its documented task boundary", "keep the output safe, reproducible, and within the documented scope at all times", "Confirm assumptions, limitations, and warnings are surfaced explicitly") plus the circular "See `## Usage` above" reference. Not 1 because the Cellosaurus-specific Usage commands are lean and genuinely useful. | 2 / 5 |
Actionability | Concrete executable commands with real examples (`get CVCL_0030 --fields id,ac,sy`, `search "id:HeLa" --rows 5`) that match the actual argparse interface in scripts/client.py. Not 5 because script paths are inconsistent across sections (`scripts/client.py` vs `.trae/skills/cellosaurus-api/scripts/client.py` vs `cd "20260316/scientific-skills/…"`), breaking copy-paste readiness, and the --format/--sort options are undocumented. | 4 / 5 |
Workflow Clarity | Sequences and validation sections exist (Recommended Workflow, run plan, Completion Checklist, Quick Validation) but are generic boilerplate untethered to the actual operations. Not 4 because the "Quick Validation" expected output ("Result file: cellosaurus_api_result.md / Validation summary: PASS/FAIL") contradicts what `python scripts/client.py --help` actually produces (argparse help text) and the script produces no result file at all. | 3 / 5 |
Progressive Disclosure | The bundle is appropriately small (two scripts, both real and referenced), but the monolithic SKILL.md carries duplicated scaffolding (two validation sections, two usage sections) and content that should simply be deleted rather than inlined, with inconsistent script paths across sections. Not 2 because headers provide real structure and no bulk content wrongly belongs in separate files. | 3 / 5 |
Total | 12 / 20 Passed |