Tech-Debt Audit

Use the Claude Code command as the canonical workflow instead of maintaining a duplicated snapshot.

Load the Canonical Command

1. Read .claude/commands/audit-tech-debt/SKILL.md completely.
2. Read .claude/commands/_audit-common.md and .claude/commands/_audit-severity.md completely before auditing.
3. Follow the canonical command's scope selection, baseline snapshot, per-dimension discovery recipes and triage, extra finding fields and TD ID convention, cross-dimension dedup, and report format.

Adapt Claude Conventions to Codex

• Treat the user's text after the skill request as the command arguments described by the canonical skill's argument-hint.
• Resolve each /audit-<name> reference to that audit's SKILL.md under .claude/commands/, and keep cross-referenced out-of-scope work (e.g. NIFAL translation correctness) as routing, not findings.
• Prefer the repository's codebase-memory graph tools for caller and consumer checks (dead code, duplication consolidation sites); use text search for the marker, stub, size, and literal greps the dimensions prescribe, plus git log/git blame aging.
• Interpret Claude Task-agent instructions as Codex sub-agent delegation only when delegation is available and allowed. Otherwise execute the routed dimensions sequentially in the canonical's impact order.
• Preserve the canonical output path and do not create GitHub issues; publishing remains a separate audit-publish action.

Do not copy the canonical command body into this skill. The Claude command is the single source of truth.