Content
56%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 has a clear workflow, a real analysis script, and mostly executable commands, but it badly over-spends tokens: it re-explains standard Python tooling Claude already knows, repeats itself across three overlapping overview sections, and inlines ~150 lines of pattern lists that duplicate references/python-patterns.md instead of deferring to it.
Suggestions
Cut the "What This Skill Analyzes" and "Automated Analysis" sections — they restate the Quick Start list and Analysis Workflow step 1 verbatim; one statement of the workflow is enough.
Move the framework-indicator lists (Django/Flask/FastAPI/CLI/Library/Data Science) and dependency-file enumerations into references/python-patterns.md, which already contains near-identical content, and replace them in SKILL.md with a one-line pointer like the existing "See: python-patterns.md" link.
Drop instruction in basic, well-known commands (venv creation, `pip install -r requirements.txt`, `flask run`, `python manage.py runserver`) — keep only the skill-specific script invocation, the output template, and any non-obvious gotchas.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | At ~330 lines the body is noticeably verbose and re-teaches what Claude already knows: venv setup (`python -m venv venv`, `source venv/bin/activate`), pip/Poetry/Pipenv/Conda install commands, `flask run` / `uvicorn main:app`, and standard framework indicators. It also duplicates itself — "What This Skill Analyzes" restates the "Analysis Workflow", "Automated Analysis" restates step 1, and the "Best Practices" section is generic advice. Not 3 because the padding is pervasive (several sections' worth), not just occasional over-explanation. | 2 / 5 |
Actionability | Provides mostly executable guidance: `python scripts/analyze_repo.py <repo_path>`, concrete venv/install/run/pytest command blocks, and a structured output template. Not 5 because several workflow steps are descriptive checklists ("Check for README", "Examine test files to understand features") rather than runnable commands, and the output template is necessarily placeholder-based. | 4 / 5 |
Workflow Clarity | A clear 6-step sequence (scan → type → entry points → dependencies → setup → functionality) is stated up front and then elaborated step-by-step, ending in a defined output format. This is read-only analysis with no destructive or batch operations, so the validation cap does not apply; the only gap is that verification of findings is left to the script's "Limitations" note rather than an explicit checkpoint. Not 5 because there is no explicit validate/verify step, and not 3 because the sequence and expected output are unambiguous. | 4 / 5 |
Progressive Disclosure | The reference "See: [python-patterns.md](references/python-patterns.md) for detailed patterns" is clearly signaled and one level deep (verified: no nested references inside it), and scripts/analyze_repo.py is properly linked. However, ~150 lines of the body (framework indicator lists, entry-point lists, dependency-file lists) duplicate content that already lives in references/python-patterns.md — content that clearly belongs in the separate file is inlined. Not 4 because the duplication between SKILL.md and the reference file is substantial, not a minor organization gap. | 3 / 5 |
Total | 13 / 20 Passed |