Content
86%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.
An excellent, highly actionable reference skill: pinned, runnable commands and code for every workflow, honest interpretation caveats, a pre-reporting validation checklist, and clean one-level-deep progressive disclosure into six real reference files. The only notable gaps are unspecific validation-feedback loops in a couple of workflows and minor trim candidates (dates, repeated uv invocations, the citation block).
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is largely lean: tight code blocks with pointed, high-value caveats (parser mismatches, normalized RF being 'undefined (None), not evidence of agreement') and no explanation of concepts Claude already knows. Minor trim candidates keep it below anchor 5: time-sensitive details in the 'Current Target' section ('released September 3, 2025 and verified ... on September 30, 2026'), six near-identical `uv run --no-project --isolated --with` invocations, and the extended citation block at the end. | 4 / 5 |
Actionability | Fully executable throughout: pinned install commands ('uv pip install "ete4==4.4.0"'), a copy-paste Quick Start, complete runnable Python snippets for every workflow (inspect/transform, RF comparison, evol events, taxonomy, visualize), and exact CLI invocations with flags ('--parser support --mode circular --show-support --color-by-support --support-scale percent'). It also covers the common cases and their concrete failure modes ('A value of 1 means 1% in the former and full support in the latter'). | 5 / 5 |
Workflow Clarity | Each workflow section is clearly sequenced with concrete inputs/outputs, and the 'Quality and Interpretation Checks' numbered checklist provides explicit pre-reporting validation (confirm parser, check duplicates, state rooting, record versions). It falls short of anchor 5 because validate→fix→retry feedback loops are not spelled out — e.g. 'A parser mismatch is the most common cause of NewickError' diagnoses but does not describe the correction loop, and taxonomy ambiguity handling raises an error without a recovery path. | 4 / 5 |
Progressive Disclosure | The body is a well-organized overview with 'Load only the reference needed for the task' and a Reference Map giving one-line descriptions of six reference files, all of which exist in references/ and are exactly one level deep (reference files link only to external upstream URLs, not to each other). Detailed material (API reference, full workflows, taxonomy setup, visualization, migration) is correctly split out of SKILL.md, matching the clear-overview anchor 5. | 5 / 5 |
Total | 18 / 20 Passed |