Content
17%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 a persona document, not an operational skill: it spends most of its length on a generic capabilities/traits/knowledge catalog that adds nothing Claude does not already know, gives abstract one-line instructions with no code, commands, examples, or validation steps, and points to a nonexistent 'resources/implementation-playbook.md' while providing no real bundle files. It needs a complete rewrite into concise, actionable guidance with concrete workflows and working references.
Suggestions
Replace the ~150 lines of Capabilities/Behavioral Traits/Knowledge Base catalog with concise actionable instructions — e.g. concrete procedures for defining KPI frameworks, building a churn model, or sizing a market, with example SQL/Python snippets or template outputs Claude can execute.
Create the missing 'resources/implementation-playbook.md' (or fix the path to an existing file) so the one reference in the body resolves; consider splitting detailed methodology per topic into one-level-deep reference files.
Turn the 'Instructions' and 'Response Approach' sections into a concrete sequenced workflow with validation checkpoints (e.g. 'define success metrics with the user before analysis; verify data quality before modeling; sanity-check model output against known benchmarks'), and replace the circular Use/Do-not-use bullets with real trigger and exclusion conditions.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Roughly 150 of the 181 lines are padded catalog material — a 'Capabilities' section listing ~70 generic bullet points ('Computer vision for image and video analytics', 'Color theory and design principles for business audiences'), plus a 'Knowledge Base' section enumerating concepts Claude already knows ('Statistical analysis and machine learning techniques', 'Cloud analytics platforms and data warehousing'). This matches anchor 2 (noticeably verbose; several unnecessary explanations or padded sections) — not anchor 1 since it is list padding rather than prose explaining basic concepts, but far above the token budget a skill should occupy. | 2 / 5 |
Actionability | There is no code, no commands, and no executable guidance anywhere. The Instructions section is entirely abstract ('Clarify goals, constraints, and required inputs. Apply relevant best practices and validate outcomes. Provide actionable steps and verification.') and the Response Approach steps are one-line abstractions ('Assess data availability and quality for analysis') with no method, tool, command, or example. This is precisely anchor 1: entirely vague or abstract; only describes rather than instructs. | 1 / 5 |
Workflow Clarity | The 'Response Approach' section lists a rough 8-step sequence, but the steps are generic one-liners with many gaps — no methodology for any step, no validation checkpoints, and no feedback loops. The 'Use this skill when / Do not use this skill when' sections are circular boilerplate ('Working on business analyst tasks'). This matches anchor 2 (rough sequence present but many gaps; steps poorly defined; validation absent) better than anchor 3, whose steps are at least concretely actionable. | 2 / 5 |
Progressive Disclosure | No bundle files exist (references/, scripts/, assets/ are absent), and the body's single pointer — 'If detailed examples are required, open `resources/implementation-playbook.md`' — references a file that does not exist, so the reference is dangling. The bulk of the content (the persona, capabilities, behavioral traits, and knowledge base) is inlined monolithic material that belongs in separate files or nowhere at all. This matches anchor 2 (minimal structure; content that clearly belongs in separate files is inlined; or references are buried) — the section headers give it slightly more structure than anchor 1, but the broken pointer prevents a higher score. | 2 / 5 |
Total | 7 / 20 Passed |