pedantic-audit

Perform a comprehensive pedantic audit of the entire repository (or the scope provided as argument).

Follow this process:

1. Discover the codebase: Use Glob and Read to understand the repository structure:

   • Identify primary language(s) used
   • Map the directory structure and module boundaries
   • Count total source files by language/extension
   • Identify configuration files, entry points, and key modules

2. Sample strategically: You cannot read every file. Sample 15-25 representative files:

   • Entry points and main modules (always include)
   • Configuration/settings files (always include)
   • 2-3 files from each major directory/module
   • Any files that define shared patterns (base classes, utilities, constants, types)
   • Prefer files that other files import from — these set conventions

3. Establish the codebase's conventions: Before flagging violations, identify what conventions the codebase already uses:

   • What casing convention is dominant?
   • How are imports organized?
   • What patterns are established for config, errors, validation?
   • What naming conventions exist for files, classes, functions?
   • Document these as "established conventions" in your report

4. For each relevant pedantic principle, invoke the corresponding pedantic-coder skill and audit the sampled files:

   • Naming Precision — scan for vague identifiers across the codebase
   • Casing Law — check for mixed conventions across files
   • Import Discipline — compare import ordering across multiple files
   • Declaration Order — check if files follow a consistent shape
   • Symmetry — compare parallel modules (services, handlers, models) for structural consistency
   • One Pattern, One Way — identify where multiple patterns exist for the same problem
   • Magic Value Elimination — grep for inline literals that should be constants
   • Dead Code Intolerance — scan for commented-out blocks, unused imports, stale TODOs

5. Check language-specific pedantry: Invoke the Python, TypeScript, or Go language pack skill as applicable.

6. Report findings in this structure:

   Codebase Overview

   • Languages, structure, total files, files sampled

   Established Conventions

   • What the codebase does consistently (acknowledge what's right)

   Findings by Principle

   For each principle with violations:

   • [Principle Name] — X violations across Y files
   • Specific examples with file:line references
   • Whether this is a codebase-wide pattern or isolated incidents
   • The fix (show before/after for representative cases)

   Codebase-Wide Patterns

   • Systemic issues that affect multiple files/modules
   • Convention conflicts (where the codebase disagrees with itself)

   Summary

   • Severity counts: Critical / Important / Nitpick
   • Top 5 highest-impact improvements (repo-wide, not per-file)
   • Overall pedantry score: pristine / clean / acceptable / messy / catastrophic
   • One-paragraph verdict

Focus on codebase-wide patterns over individual file issues. The goal is to identify systemic inconsistencies and establish a path to a pristine codebase.