Content
19%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill reads like a product specification document rather than an actionable skill for Claude. It is extremely verbose (~600+ lines) with extensive placeholder templates, repetitive pseudo-specifications, and six lengthy output examples that could be drastically condensed. The content lacks any executable code despite being a data analysis skill, and everything is crammed into a single monolithic file with no progressive disclosure structure.
Suggestions
Reduce the output format template to a brief structural outline (10-15 lines) and move the full template to a separate REPORT_TEMPLATE.md file
Replace pseudocode bullet lists in analysis steps with actual executable Python/code snippets for computing correlations, risk scores, and trend analysis
Consolidate the 6 usage examples down to 1-2 brief examples, or move them to a separate EXAMPLES.md file
Extract the detailed risk scoring algorithm, report template, and data schema references into separate bundle files, keeping SKILL.md as a concise overview with clear navigation links
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Extremely verbose at 500+ lines. Extensively explains concepts Claude already knows (what PHQ-9 is, how correlation works, what emotions are). The output format template alone is ~200 lines of placeholder text. The report structure, examples, and step descriptions are massively padded with repetitive pseudo-specifications that read more like a product requirements document than actionable skill instructions. | 1 / 5 |
Actionability | Despite its length, the skill contains zero executable code. All 'analysis' steps are bullet-point descriptions of what to analyze rather than how to compute it. The code blocks contain pseudocode lists (e.g., 'calculate correlation coefficient') rather than actual implementations. The risk scoring algorithm is the most concrete element, but even that is presented as a specification rather than executable logic. | 2 / 5 |
Workflow Clarity | The 10-step sequence is clearly numbered and logically ordered, with the crisis risk assessment appropriately marked as highest priority. However, there are no validation checkpoints between steps, no feedback loops for error recovery during analysis, and the data validation in Step 1 is vague ('check file exists, verify structure'). The crisis detection workflow is the strongest part with clear thresholds and action triggers. | 3 / 5 |
Progressive Disclosure | Monolithic wall of text with no references to external files for detailed content. The entire report template, all examples (6 lengthy ones), all analysis specifications, and the output format are inlined in a single massive document. No bundle files exist to offload content to. The report template alone should be a separate file, and the detailed scoring algorithms could be in a reference document. | 1 / 5 |
Total | 7 / 20 Passed |