qredence/fleet-optimize

Optimize prompts, code, configurations, schemas, SQL, regexes, and other scorable text artifacts with the latest upstream GEPA parallel-proposal workflow. Use when a candidate should be improved by P×N sampling, concurrent evaluation, Pareto-aware selection, or grouped batch evaluation.

Improve scorable prompts, programs, configs, specs, SQL, schemas, regexes, and agent instructions with GEPA Omni. Use when a candidate can be scored by an objective evaluator or useful LLM-as-judge feedback.