Content
61%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.
A highly actionable, well-structured body with real bundle files and clear navigation. Main weaknesses are redundant/duplicated sections, a few code examples referencing undefined variables, and missing validation checkpoints in batch/destructive tree-processing workflows.
Suggestions
Add explicit validation steps to batch/destructive workflows (e.g. verify leaf counts or topology before/after pruning and node deletion in Use Case 4, and confirm output files were written).
Remove duplicated content: the Qt dependency install block, NCBI setup instructions, and the placeholder 'Load these references' comment block.
Define or inline the undefined variables ('extract_species_from_name', 'species_to_taxid') so every code example is copy-paste executable, and move the longer per-capability examples into references/workflows.md to slim the SKILL.md body.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly executable code (efficient), but contains avoidable redundancy: Qt system-dependency install instructions appear twice, NCBI database setup appears twice, and the 'Load these references when detailed information is needed' pseudo-Python comment block is pure padding that adds no information beyond the reference list above it. | 3 / 5 |
Actionability | Concrete, executable ete3 code and real CLI commands for the bundled scripts cover the common cases well. Minor gaps: 'extract_species_from_name' in the taxonomy annotation example and 'species_to_taxid' in Use Case 1 are referenced but never defined, so those snippets are not copy-paste ready. | 4 / 5 |
Workflow Clarity | Use Case 1 presents a clean numbered 1-6 pipeline, but workflows involving batch and destructive operations lack validation checkpoints: Use Case 4 iterates a directory of trees and deletes low-support nodes with no verification step before writing outputs, which caps workflow clarity at 3 per the rubric guidelines. | 3 / 5 |
Progressive Disclosure | Three real reference files (api_reference.md, workflows.md, visualization.md) and two real scripts, each clearly described with its contents and one level deep. However, the body itself carries multiple full code examples per capability plus four complete use cases, some of which duplicates material that belongs in workflows.md, so the split could be leaner. | 4 / 5 |
Total | 14 / 20 Passed |