darwinian-evolver

Evolutionary search optimizes LLM prompts and SQL queries by running an iterative, LLM-driven loop that refines candidates against a defined fitness function. You define a Problem with an Organism, Evaluator, and Mutator to guide the evolutionary process toward optimal results.

To run an LLM-driven evolutionary search loop for code optimization, you need Python 3.11, git, uv, and one of the following API keys: OPENROUTER_API_KEY, ANTHROPIC_API_KEY, or OPENAI_API_KEY to execute the optimization process.

Yes, you can use evolutionary search to optimize regex patterns for filtering datasets. The darwinian-evolver skill specifically supports evolving regular expressions to effectively filter out unwanted entries from your dataset through iterative refinement.

The best way to iteratively refine code snippets against a fitness function is using an LLM-driven evolutionary search loop. This approach targets optimization tasks by defining a Problem with an Organism, Evaluator, and Mutator to guide continuous improvement.

You should use an evolutionary search loop instead of manual prompt engineering when your optimization task requires iterative refinement against a specific fitness function. It automates the evolution of prompts, regex, SQL, and code, which is inefficient to refine manually.