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 body delivers concrete, executable commands and a usable parameter/config reference, but it is padded with repeated Cloudflare warnings and generic boilerplate, and it inlines large API/config/algorithm blocks that overlap with references/README.md rather than pointing out to them. Workflow steps are templated rather than skill-specific.
Suggestions
Collapse the repeated Cloudflare/bioRxiv-medRxiv caveats into a single Known Issues section and reference it once from Usage/Examples.
Move the full API endpoint reference, config.yaml, and output-format samples into references/ (they already overlap references/README.md) and keep only a minimal quick-start example in SKILL.md.
Replace the generic templated Workflow/Output Requirements/Error Handling sections with the skill's actual flow: install deps → choose source → run main.py → review output, with the py_compile/smoke_test checks inline as checkpoints.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is noticeably verbose: the bioRxiv/medRxiv Cloudflare caveat is repeated across Overview, Network Access Notice, Usage, Examples, and Known Issues, and several sections (When to Use, Key Features, Implementation Details, Workflow, Output Requirements, Error Handling, Response Template) are generic boilerplate Claude already knows. It is above a 1 because the command/parameter material is accurate and useful, but the padding and repetition place it below the mostly-efficient 3. | 2 / 5 |
Actionability | Concrete executable commands (`python scripts/main.py --sources arxiv --days 7 --output json`), a full parameter table, a config.yaml example, and JSON/markdown output samples give mostly copy-paste-ready guidance. It is below 5 because the installation block hardcodes a personal path (`/Users/z04030865/.openclaw/workspace/...`) and some example blocks are missing closing fence context, leaving minor gaps. | 4 / 5 |
Workflow Clarity | A sequenced Workflow list and explicit validation commands (py_compile Quick Check, smoke_test Audit-Ready) exist, so it is above a 2. However the numbered workflow is generic templated steps rather than this skill's actual install→configure→run→review flow, and the validation checkpoints sit in separate sections rather than inline in the sequence, matching the anchor with steps present but checkpoints implicit/fragmented. | 3 / 5 |
Progressive Disclosure | Bundle files exist (references/README.md, scripts/main.py, smoke_test.py) and the body signals them, but large blocks that belong in references are inlined in SKILL.md — full API endpoint docs, the complete config.yaml, output-format samples, and the trending algorithm — and the API material is duplicated in references/README.md. This fits the anchor with some structure and signaled references but content that should be separate is inline, short of the well-split 4. | 3 / 5 |
Total | 12 / 20 Passed |