ralph
Ralph is an autonomous AI agent loop that runs repeatedly until all PRD items are complete.
Ralph is an autonomous AI agent loop designed to drive AI coding tools like Amp or Claude Code through iterative development cycles until all items in a Product Requirements Document (PRD) are complete. Each iteration spawns a fresh AI instance with clean context, maintaining memory through git history, progress.txt, and prd.json files. The agent autonomously selects the highest-priority incomplete user story, implements it, runs quality checks, commits successful changes, and repeats until all tasks pass or the iteration limit is reached. Ralph includes companion skills for generating structured PRDs and converting them to the JSON format the agent loop consumes.
Key Features
Use Cases
- 01Autonomously implementing multi-story feature branches from detailed PRDs without manual intervention
- 02Breaking down large development tasks into context-window-sized stories that AI can complete reliably
- 03Building up institutional knowledge through AGENTS.md files that persist learnings across iterations
- 04Running overnight development cycles that implement, test, and commit multiple user stories
- 05Maintaining quality standards through automated typecheck and test verification loops
- 06Handling frontend development with automated browser verification of UI changes
Related Agents
View morehermes-agent
The agent that grows with you
agency-agents
A complete AI agency at your fingertips - From frontend wizards to Reddit community ninjas, from whimsy injectors to reality checkers. Each agent is a specialized expert with personality, processes, and proven deliverables.
openinterpreter
A coding agent for open models like Kimi K3
cline
Autonomous coding agent as an SDK, IDE extension, or CLI assistant.
ralph — FAQ
What is Ralph and how does it work?+
Ralph is an autonomous AI agent loop that repeatedly runs AI coding tools (Amp or Claude Code) to implement items from a Product Requirements Document. Each iteration picks an incomplete story, implements it, runs quality checks, commits if tests pass, and repeats until all stories are complete or the iteration limit is reached.
How do I install Ralph in my project?+
You can copy the ralph.sh script and prompt templates into your project's scripts/ralph/ directory, install the skills globally to your Amp or Claude config directories, or add it as a Claude Code marketplace plugin with /plugin marketplace add snarktank/ralph. You'll need either Amp CLI or Claude Code already installed and authenticated.
What are the prerequisites for running Ralph?+
You need either Amp CLI or Claude Code (via npm install -g @anthropic-ai/claude-code) installed and authenticated, jq installed (brew install jq on macOS), and a git repository for your project. No API keys are required beyond what your chosen AI coding tool already uses.
Which AI coding clients does Ralph work with?+
Ralph works with Amp CLI (default) and Claude Code. You select the tool using the --tool flag when running ralph.sh, and it uses the appropriate prompt template (prompt.md for Amp or CLAUDE.md for Claude Code).
Is Ralph free to use?+
Yes, Ralph itself is open source and free. However, you'll incur costs from the underlying AI coding tool you choose (Amp or Claude Code), which consume API credits during each iteration.
How do I create a PRD for Ralph to execute?+
Use the included /prd skill by asking your AI tool to "create a prd for [feature]" which generates a detailed markdown requirements document. Then use the /ralph skill to convert that PRD to the prd.json format that the autonomous loop consumes.
How do I install ralph?+
Open the source repository on GitHub and follow its README. ralph is a agent — MCP Agents Market links you directly to the official repo.
Is ralph free?+
ralph is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.