Run GEPA on your favorite non-python frameworks.

GEPA reflective prompt evolution toolkit — Pareto-rank prompt candidates, score with pluggable LLM judges, persist trials.

TypeScript implementation of GEPA (Gradient-free Evolution of Prompts and Agents) - Complete port with 100% feature parity

GEPA (Genetic-Pareto) optimizer for TypeScript - Multi-objective optimization through reflective evolution, with first-class support for AI SDK's generateObject

No description provided.

Eval framework for Noetic agents: the describe/it/scorer suite runner, built-in and custom scorers, regression baselines, and GEPA-backed prompt optimization.

Genetic-Pareto prompt optimizer to evolve system prompts from a few rollouts with modular support and intelligent crossover

Skill scanner and quality scorer for Pi. Scans, scores, and identifies weak skills. Pairs with pi-evolve for mutation.

React components for visualizing GEPA prompt-optimization runs.

GEPA-style prompt self-optimization for AI agents. Prompts evolve from execution traces — auto-mutate, measure, rollback bad changes. Zero LLM calls.

The Dak Stack deepen generator — the self-improving loop's ORIGINATOR (the eval gate is the validator; this proposes what the gate judges). proposeRevision is a GEPA/DSPy-style hill-climb over a skill's natural-language scaffolding (the SKILL.md body): a

Local code search for AI agents: six fast, purpose-built tools that return ranked answers, not raw grep. Because maybe grep isn't all you need... 🍬

DSPyGround - Optimize and test your AI agents with AI SDK

LangGraph.js adapter for darwin-agents — wrap self-evolving Darwin agents as StateGraph nodes with zero hard deps. V0.5.4 opens the darwin-agents peer range to any 0.x (ends the per-minor install breakage; verified against 0.11). V0.5.2 wires the runAgent

Evolutionary self-improvement for Pi. AlphaEvolve-inspired mutation, evaluation, and selection of prompts, skills, and extensions. Open-ended agent evolution.

Self-improving agent for Pi with memory and skills

Effect-native typed model programs, immutable compilation artifacts, bounded optimization, and receipted prediction.

Pi extension for project-local learned context priors.

CLI and skills for building repo-local agent self-improvement systems.

Tracing and evaluation utilities for TypeScript LLM applications.