grandamenium

Evolve Agent — Testing skill for Claude Code

Testing community

Build agents that improve their own prompts through automated iteration and testing.

How to install Evolve Agent

This entry records only its repository, not the path inside it, so there is no exact command to give. Open grandamenium/evolve-agent and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Evolve Agent does

Build agents that improve their own prompts through automated iteration and testing.

Alternatives in Testing

Deslop Reborn — One iteration of the IronClaw Reborn de-slop loop — take ONE Reborn crate, fan out parallel review sub-agents 12.6k ★
Darwin Skill — 达尔文.skill —— 一个让你的Skill无限进化的系统：评估→改进→测试→保留或回滚 Autoresearch-inspired autonomous skill optimization for Claude C 5.7k ★
Claude Code Spec Workflow — Automated Kiro-style Spec workflow for Claude Code 3.6k ★

README

evolve-agent

Build agents that improve their own prompts through automated iteration and testing. Instead of manually tweaking prompts, let your agent learn from failures and refine itself.

Quick Start

# Install the skill
git clone https://github.com//evolve-agent ~/.claude/skills/evolve-agent
cd ~/.claude/skills/evolve-agent
npm install

# Run it
npx evolve-agent --agent test-agent.js --task-file test-cases.js --iterations 3

What It Does

Baseline Testing — Runs your agent N times against test cases, measures accuracy, speed, and cost
Failure Analysis — Collects all failures and identifies patterns
Prompt Refinement — Uses Claude to generate improved prompts based on failure patterns
A/B Testing — Tests the refined prompt and compares to baseline
Auto-Accept — If improved, saves the new prompt; if not, reverts
Iteration — Repeats for N cycles until improvements plateau

Requirements

Node.js 18+
npm 9+
@anthropic-ai/sdk (installed by npm)
ANTHROPIC_API_KEY environment variable set

Installation

git clone https://github.com//evolve-agent ~/.claude/skills/evolve-agent
cd ~/.claude/skills/evolve-agent
npm install

Usage

Basic Usage

npx evolve-agent \
  --agent my-agent.js \
  --task-file test-cases.js \
  --iterations 3 \
  --runs-per-iteration 5

All Options

npx evolve-agent \
  --agent                     # Agent file (required)
  --task-file                 # Test cases file (required)
  --iterations 3                     # Evolution cycles (default: 3)
  --runs-per-iteration 5             # Runs per cycle (default: 5)
  --improvement-threshold 10         # Min % improvement to accept (default: 10)
  --output-dir ./results             # Output directory (default: current dir)
  --model claude-3-5-sonnet-20241022 # Claude model (default: sonnet)
  --evaluator custom-eval.js         # Custom evaluator (opti