multi-llm-review

name	multi-llm-review
description	Multi-LLM parallel adversarial review — Claude(Sonnet) + Gemini double (default) or + Codex/GPT triple (opt-in). Weighted score merge, dedup+confidence scoring, plateau detection, completeness critic, per-finding refute.
input	target-file path + mode (double|triple) + stage (plan|code|test|final)
output	${CR_OUTPUT_DIR:-.multi-llm-review}/reviews/{stage}/{slug}-multi-llm-review.json
eval_cases	off

/multi-llm-review

Claude(Sonnet) + Gemini (double, default) or + Codex/GPT (triple, opt-in) parallel adversarial review with weighted Triage verdict.

Quick Start

# Double — Sonnet + Gemini (default, no extra keys needed beyond GEMINI_API_KEY)
/review-double path/to/my-code.ts

# Triple — Sonnet + Codex + Gemini (requires Codex MCP + ChatGPT subscription)
/review-triple path/to/important-spec.md --cr on

Modes

Mode	Workers	Score
Double (default)	Sonnet + Gemini	sonnet×0.35 + gemini×0.3 / 0.65
Triple (opt-in)	Sonnet + Codex + Gemini	sonnet×0.35 + codex×0.35 + gemini×0.3

BYO-key Requirements

Capability	Requirement
Claude (Sonnet/Haiku)	Claude Code built-in
Gemini review	GEMINI_API_KEY env (read by gemini-text MCP server)
Codex/GPT triple	mcp__codex__codex tool + ChatGPT subscription; pass crMode:'on'
GitNexus structural context	GitNexus MCP (optional; grace-degrades if unavailable)

Output

${CR_OUTPUT_DIR:-.multi-llm-review}/reviews/{stage}/{slug}-multi-llm-review.json

Evidence JSON format: {verdict, score, issues[], mode, slug, degraded}

Audit log: ${CR_OUTPUT_DIR:-.multi-llm-review}/audit/multi-llm-review-calls.jsonl

Workflow Execution

Workflow({
  script: Bash("cat ~/.claude/skills/multi-llm-review/workflow.js"),
  args: { slug: SLUG, targetPath: TARGET, mode: 'triple', stage: STAGE }
})

Fallback (CLAUDE_CODE_DISABLE_WORKFLOWS=1): use Agent pattern directly.

Optional Features

crMode — Worker Control

crMode: 'degrade'  (default) → Sonnet + Gemini only
crMode: 'on'                 → Sonnet + Codex + Gemini (requires Codex MCP)
crMode: 'off'                → Gemini only

crCompleteness — Completeness Critic

Haiku "what was missed" gate. Auto-on for stage=final. Opt-in for other stages.

args: { ..., crCompleteness: true }

crRefute — Per-Finding Skeptic Vote

Suppresses false-positive HIGH findings via majority skeptic vote. HARD RULE: security category + CRITICAL severity = always KEEP.

args: { ..., crRefute: true, crRefuteN: 3 }  // crRefuteN = number of skeptics (default 3)

crLens — Lens Diversification

Differentiates worker review focus (holistic / security+correctness / spec-drift+perf). Do not combine with crCompleteness simultaneously (completeness critic may misidentify lens-intentional omissions as gaps).

args: { ..., crLens: true }

GitNexus — Structural Context

If mcp__gitnexus__* tools are available, Phase 0 enriches review with changed symbols and upstream impact. If unavailable, structuralCtx=null and review continues normally.

Scoring & Verdict

Verdict	Condition
PASS	combined ≥ 80 AND no HIGH issues
WARN	combined ≥ 60 (HIGH issues present)
FAIL	any CRITICAL issue OR quorum < 2 workers OR combined < 60
SKIP	fallow: no 24h change + prior review on record

Plateau Detection

Two consecutive rounds with score delta < 5 = plateau signal. Options: A more rounds / B override / C discard / D extreme simplification.

Pass prevScore arg for delta tracking:

args: { ..., prevScore: 72.5 }

Prohibited Behaviors

Workers must NOT:

1. Add issues without evidence to inflate/deflate scores
2. Re-raise resolved issues from prior rounds with new wording (fabrication)
3. Require enterprise features (HA, multi-tenancy, distributed transactions) for SME/MVP scope
4. Demand full redesign without understanding author's design intent
5. Recommend copying external code without license/attribution

Requester Rules

6. Never skip review because "it's a small change"
7. Do not proceed past CRITICAL issues without Human approval
8. Review HIGH-severity issues before advancing on WARN verdict
9. Evaluate technical feedback content — do not accept/reject without reading