StackMap
Subscribe

AutoGen vs council-of-high-intelligence

Multi-agent conversation framework for building LLM applications with cooperating agents. — versus — /council: 18 AI personas deliberate your hardest decisions across multiple LLM providers — structured multi-round disagreement, confidence-weighted verdicts, one slash command.

The curated verdict

Both stage multi-agent deliberation; AutoGen is the framework you build conversations with, Council is the finished product — personas, protocol and verdict included, one command away.

AutoGencouncil-of-high-intelligence
Stars60k3.6k
Forks9.0k320
LanguagePythonShell
LicenseCC-BY-4.0MIT
Last activity3 months ago2 days ago
Topicsagents, evalsagents, skills
Curated connections92

AutoGen — the curator's take

Multi-agent conversation framework for building LLM applications with cooperating agents.

council-of-high-intelligence — the curator's take

Structured disagreement as a product: 18 personas with genuinely different priors argue your decision across providers (real model diversity, not one model roleplaying), through quick/standard/deep deliberation modes, ending in a confidence-weighted verdict. Its own README has a 'When Not to Use It' section — our kind of project. Best for irreversible, ambiguous calls where you'd otherwise ask three friends. NOT for anything with a checkable answer (deliberation theater costs real tokens), and a council is still only as wise as its training data — it widens perspective, it doesn't add ground truth.