llm-debate

Claude Code skill: Multi-agent adversarial review panel — 4-6 AI reviewers debate your code/plans, then a supreme judge delivers the verdict. 9 auto-detected signal groups, built-in domain checklists, anti-groupthink mechanisms.

Multi-agent deliberation system for Claude Code - 3-vendor heterogeneous ensemble (Claude + Gemini + OpenAI)

Roundtable — Put multiple AI models in a room, give them personas, and watch them fight for consensus. The ultimate multi-AI debate engine.

Structured refinement of any document between AI agents. Two agent roles bounce a file back and forth using disagreement markers until it converges. Open source, agent-agnostic, runs from any terminal.

Claude Code plugin for second opinions: iterative adversarial debates between Claude and Codex over any subject, from specs, designs, and code to non-technical positions, ending in genuine agreement or a documented dispute.

Models debating in Cursor

Structured cross-model peer review for DeepSeek Harness — blind reasoning, anonymous critique, revision, and deterministic verdicts.

Multi-model debate coordinator with participant selection supporting Gemini, OpenAI, Claude, DeepSeek, and Grok.

An intelligent multi-agent AI debate arena and consensus engine with adversarial cross-examination, AI judge synthesis, and rich visual telemetry.

Ask one question, several models answer, and it shows you exactly where they disagree. A local-first council of Ollama and cloud models with debate rounds, a consensus gate, grounding checks and blind picks.