style-audit

name: style-audit description: Evaluate style, readability, and documentation quality with evidence-backed findings and fix guidance. allowed-tools:

• Read
• Write
• Edit
• Bash
• Glob
• Grep
• Task
• TodoWrite model: sonnet x-version: 3.2.0 x-category: quality x-vcl-compliance: v3.1.1 x-cognitive-frames:
• HON
• MOR
• COM
• CLS
• EVD
• ASP
• SPC

STANDARD OPERATING PROCEDURE

Purpose

Assess code style, naming, organization, and documentation to raise maintainability while respecting project conventions.

Trigger Conditions

• Positive: pre-merge polish, refactoring support, documentation sweeps, or readability reviews.
• Negative: functionality debugging (use functionality-audit) or architectural coupling checks (use connascence-quality-gate).

Guardrails

• Confidence ceiling: Add Confidence: X.XX (ceiling: TYPE Y.YY) with ceilings {inference/report 0.70, research 0.85, observation/definition 0.95}.
• Evidence requirements: Provide file:line, rule/reference (style guide, lint rule), and example fix.
• Structure-first: Keep examples/ and tests/ updated with preferred wording and format for findings.
• Local conventions: Respect project-specific standards; flag deviations separately from general best practices.

Execution Phases

1. Context & Standards
   • Identify applicable style guides, lint rules, and documentation standards.
   • Note exceptions or overrides for the repository.
2. Review Passes
   • Naming & API clarity; comments and docs alignment.
   • Structure & readability (function size, duplication, complexity hints).
   • Documentation & developer experience (README, usage examples).
3. Findings & Fixes
   • Record issues with file:line, rule reference, and suggested rewrite or refactor.
   • Group by severity (blocker/major/minor/nit) and theme.
4. Validation & Confidence
   • Spot-check for false positives; ensure fixes do not change behavior.
   • Provide residual risks and confidence with ceiling.

Output Format

• Scope and standards applied.
• Findings list with severity, evidence, and recommended fixes.
• Quick wins vs. structural refactors separated.
• Confidence statement using ceiling syntax.

Validation Checklist

• Applicable standards identified; exceptions noted.
• Findings include file:line, rule reference, and fix guidance.
• Severity and behavior impact clarified.
• Confidence ceiling provided; output in English.

Confidence: 0.71 (ceiling: inference 0.70) - SOP rewritten per Prompt Architect confidence discipline and Skill Forge structure-first documentation.