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 well-structured with concrete, executable examples and a clean reference layout, but it is weighed down by generic templated boilerplate and a Quick Validation / Output Contract that do not match the actual script. Tightening the boilerplate and correcting the validation contract would lift the weaker dimensions.
Suggestions
Remove or replace the generic boilerplate sections (When Not to Use, Required Inputs, Recommended Workflow, Output Contract, Validation and Safety Rules, Failure Handling) with bioRxiv-specific guidance; they restate what Claude already knows and inflate the context.
Fix the Quick Validation section: `scripts/biorxiv_search.py --help` is not implemented (the script parses raw sys.argv with no argparse), so document the real CLI form `python scripts/biorxiv_search.py <keyword> [days_back]`.
Align the Output Contract with the script's actual behavior — it prints JSON to stdout rather than producing a `biorxiv_database_result.md` file with a validation summary.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Core sections (Example Usage, Implementation Details) are efficient, but a large run of generic boilerplate (When Not to Use, Required Inputs, Recommended Workflow, Output Contract, Validation and Safety Rules, Failure Handling) restates guidance Claude already knows and could be tightened, matching the score-3 anchor. | 3 / 5 |
Actionability | Provides copy-paste-ready Python and CLI examples that are mostly executable, but the Quick Validation `--help` command and the `biorxiv_database_result.md` output contract do not match the actual script behavior, leaving minor gaps. | 4 / 5 |
Workflow Clarity | The end-to-end Python example sequences search, metadata print, and PDF download clearly; however the "Recommended Workflow" steps are generic and the validation path is inconsistent with the implementation, so it falls just below a 5. | 4 / 5 |
Progressive Disclosure | Content is well organized into signaled sections with a real one-level-deep reference to `references/api_reference.md` and the `scripts/biorxiv_search.py` entry point, with only minor organization gaps from the inlined boilerplate. | 4 / 5 |
Total | 15 / 20 Passed |