Copilot Instructions

# Agentic Code Auditor

When asked to audit, evaluate, or assess this codebase for AI/LLM readiness, follow the code audit workflow defined in the audit instruction file.

## Quick Reference

This project uses an AI readiness audit system with 63 rules across 11 dimensions:
1. File Size — max file/function/class lengths, complexity, nesting
2. Naming — descriptive names, consistency, no generics/abbreviations
3. Comments — WHY not WHAT, no stale comments, appropriate density
4. Structure — single responsibility, no circular deps, clean interfaces
5. Tests — coverage >=80%, test existence, descriptive names, isolation
6. Context Files — AI context files exist with required sections
7. Safety — no secrets, proper error handling, pinned deps, no dead code
8. Types — type annotations, strict mode, no `any` casts
9. Language-Specific — per-language best practices
10. Tooling — linter, formatter, CI/pre-commit configured
11. Repo Hygiene — README, gitignore, generated file markers

Scoring: critical=10, high=5, medium=2, low=1 severity weights. WARN results receive 50% credit.
Grade: A+ >=95, A 90-94, B+ 85-89, B 80-84, C+ 75-79, C 70-74, D 60-69, F <60.

For the full audit workflow and rule details, see the code-audit instruction file in `.github/instructions/`.

Rule files are in `.audit-rules/`:
- #file:.audit-rules/index.md — full rule index
- #file:.audit-rules/scoring.md — scoring framework
- Dimension files: `01-file-size.md` through `11-repo-hygiene.md`
- Language files: `09-lang-python.md`, `09-lang-typescript.md`, `09-lang-go.md`, `09-lang-rust.md`, `09-lang-java.md`
- Custom rules: `.audit-rules/custom/*.md`
- Fix templates: `.audit-rules/../fixes/` (context-file, linter-config, formatter-config, typescript-strict, gitignore, precommit-ci)
- Example report: `.audit-rules/../fixes/example-report.md`

Supported audit modes: full (default), quick (critical only), differential (changed files), dimension-specific.