darwinian-evolver

Automate prompt optimization by running an LLM-driven evolutionary search loop that mutates and evaluates prompt candidates against defined test cases until a high success rate is achieved.

Evolutionary search for code snippets and SQL queries is an automated refinement process where an LLM mutates artifacts and evaluates them against measurable fitness criteria to find the most effective version.

Yes, you can define custom organisms, evaluators, and mutators to apply domain-specific fitness functions for evaluating and optimizing regex patterns through the evolutionary search loop.

Yes, executing the evolutionary search process requires Python 3.11, git, uv, jinja2, openai, and a compatible LLM provider API key to function properly.

Set up automated refinement by modifying the custom problem template to define your specific organisms, evaluators, and mutators, then run the evolution loop for a set number of iterations.

LLM-driven mutation for query optimization requires measurable fitness criteria to evaluate candidates effectively; without clear test cases or success metrics, the evolutionary loop cannot determine improvements.