do-competitively

Execute tasks through competitive multi-agent generation, meta-judge evaluation specification, multi-judge evaluation, and evidence-based synthesis

Independent GitHub snapshot · Automated classification · Original documentation governs