# Consensus Loop:audit

> Run a consensus-loop audit manually — Codex reviews pending trigger_tag items in the watch file. Use when you want to trigger an audit without editing the watch file, re-run a failed audit, or test the audit prompt.

- Skill: `berrzebb/consensus-loop-audit` (Agent Skill)
- Install (CLI): `npx skillmds@latest add berrzebb/consensus-loop-audit`
- Raw SKILL.md: https://api.skillmd.com/api/skills/berrzebb/consensus-loop-audit/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: berrzebb (https://skillmd.com/u/berrzebb)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/berrzebb/consensus-loop-audit

---


# Manual Audit

Run the audit process manually.

## Execute

```bash
node ${CLAUDE_PLUGIN_ROOT}/audit.mjs {{ arguments }}
```

## After Completion

Read the respond file (gpt.md) and summarize:
- Verdict per item (agree_tag or pending_tag)
- Rejection codes with reasons
- Recommended next steps

## Options

| Flag | Effect |
|------|--------|
| `--dry-run` | Print audit prompt without executing |
| `--no-resume` | Fresh session (don't resume previous) |
| `--auto-fix` | Auto-correct via Claude CLI after audit |
| `--model <name>` | Model override (default: gpt-5.4) |
| `--reset-session` | Delete saved session before running |
| `--watch-file <path>` | Override watch file path (worktree support) |