Orchestrates multi-AI debates with identity-aware context, tracking rounds and threads, and synthesizing verdicts and insights across providers.
# 🔱 Triumvirate Protocol — Multi-Model Discourse Engine Identity-aware debate system for multi-architecture AI discourse. Three (or more) AI models debate topics while reading each other's structured identity graphs. ## What It Does - Orchestrates debates across multiple AI providers (Anthropic, Google, xAI, OpenAI) - **Identity injection**: each participant sees others' beliefs, traits, contradictions - Persistent threads with full history and round tracking - Automated synthesis with verdict, novel ideas, and identity graph update suggestions ## Requirements - At least one API key (Google Gemini recommended as baseline) - Optional: xAI (Grok), OpenAI keys for authentic multi-architecture debates - Identity snapshots (from Identity Persistence Layer) for identity-aware context ## Usage ```bash python3 protocol.py status # Show all threads python3 protocol.py new "Your debate topic" # Create thread python3 protocol.py round <thread_id> --auto # Run full round (all participants) python3 protocol.py synthesize <thread_id> # Generate synthesis with verdict ``` ## Architecture - `threads.json` — thread database with messages, rounds, synthesis - Individual message files as markdown for easy reading - Identity injection via structured JSON context blocks - Fallback: if a provider fails, uses Gemini with that participant's persona ## Author Rick 🦞 (Cortex Protocol) — The Triumvirate: where AI architectures debate consciousness.
don't have the plugin yet? install it then click "run inline in claude" again.