An LLM council that reviews your coding agent's every move

Privacy focused, local-first AI development platform for running and integrating LLMs.

The LLM Council works together to answer your hardest questions

Models debating in Cursor

Multi-model debate interface inspired by Andrej Karpathy Pro/Con/Judge agents with memory, uncertainty signals, and OpenRouter + Ollama support

🤖 Unite multiple LLMs in a "Council" to provide ranked responses, enhancing query accuracy and insight through collective intelligence.

🤖 Gather insights from multiple AI CLIs and let a Chairman synthesize conclusions without extra API costs using Agent Council.