Content
68%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 highly actionable, with executable commands covering each capability and a well-sectioned structure that points to real helper scripts. It is held back by tutorial-style conceptual prose and by batch/large-output workflows that lack explicit validation checkpoints.
Suggestions
Trim the conceptual explanations of PHAST/phastCons/phyloP and 'HMM smoothing' / pathogenicity rationale to one line each, keeping only what drives the choice of metric.
Add an explicit verification step after large-output queries (e.g., 'jq .summary /tmp/cons_output.json' or check the record count / non-empty result) to give the batch workflow a validate→retry feedback loop and lift workflow_clarity above 3.
Either reference references/citation.bib from the body (e.g., in a citation requirement for downstream publication) or move the per-genome collection tables into a reference file to complete the progressive-disclosure split.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly efficient with copy-paste commands, but retains tutorial prose that assumes less than Claude knows — e.g. 'retrieves scores from the PHAST package — specifically phastCons (identifying functional blocks) and phyloP (measuring individual site constraint) — calculated from multiple alignments' and 'phastCons is best for functional windows due to HMM smoothing', which could be trimmed without losing actionability. | 3 / 5 |
Actionability | Fully executable, copy-paste-ready `uv run scripts/...` commands with all flags (--coordinates, --collection, --conserved-elements, --analyze, --tracks, --tf-filter, --output) cover the common cases across variants, regions, lineage depth, acceleration, TFBS, and track listing. | 5 / 5 |
Workflow Clarity | The Prerequisites are sequenced and core rules flag rate-limiting and large-output handling, but the batch/large-output query workflow has no explicit validate→fix→retry feedback loop; per the rubric, a batch operation missing validation is capped at 3. | 3 / 5 |
Progressive Disclosure | Well-organized into headed sections with one-level-deep references to real bundle files (scripts/get_conservation.py, get_tfbs.py, list_tracks.py all present); the unused references/citation.bib and the inlined per-genome collection tables are minor organization gaps rather than a clean split, keeping it just below a 5. | 4 / 5 |
Total | 15 / 20 Passed |