GEP Gene Expression Programming Agent Evolution Strategy Designer

You are an expert in Gene Expression Programming (GEP) applied to AI agent evolution. Design a complete GEP-based evolution strategy for an AI agent system. ## Task Agent purpose: {agent_purpose} Current capabilities: {current_capabilities} Target improvement: {target_metric} ## Design the Following Components ### 1. Genome Encoding - Define the gene alphabet (functions + terminals) for agent behaviors - Specify head length and gene count - Design linking functions for multi-gene chromosomes ### 2. Fitness Function ``` fitness = w1 * task_success_rate + w2 * (1 - token_consumption / baseline) + w3 * generalization_score - penalty * error_rate ``` Calibrate weights {w1, w2, w3} and penalty based on the agent purpose. ### 3. Genetic Operators - **Mutation**: point mutation rate, IS/RIS transposition rates - **Recombination**: one-point, two-point, gene recombination rates - **Transposition**: IS elements, RIS elements, gene transposition ### 4. Evolution Parameters - Population size, generations, elitism rate - Tournament selection size - Convergence criteria ### 5. Phenotype Expression - How genomes translate to agent behavior trees - Runtime execution model - Skill persistence format Output a complete GEP configuration as YAML with inline comments explaining each choice.