# michaelshimeles/ralphy

My Ralph Wiggum setup, an autonomous bash script that runs Claude Code, Codex, OpenCode, Cursor agent, Qwen & Droid in a loop until your PRD is complete.

Repository: https://github.com/michaelshimeles/ralphy
Canonical: https://ross.abutalabs.com/products/ralphy
Homepage: https://ralphy.goshen.fyi
Language: TypeScript
License Family: other
Topics: claude-code, opencode, ralph-loop, ralph-wiggum, ralphy
Last push: 2026-02-05T16:28:20+00:00

## Health v2 (maintenance only)
Score: 45/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 66, release rhythm 35, longevity 16
- inputs: {"age_days": 230, "days_push": 209, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2964, forks 367 (observed 2026-08-28T04:07:32.445529+00:00)

## What it is
Ralphy is a CLI tool (npm package and bash script) that runs AI coding agents like Claude Code, Codex, OpenCode, Cursor, Qwen-Code, and Factory Droid in an autonomous loop until tasks from a PRD, task list, or GitHub issues are complete. It manages git branches per task, auto-commits changes, creates PRs, and supports parallel execution via git worktrees.

## Use cases
- run an AI agent in a loop until my PRD is complete
- automate fixing a list of bugs with Claude Code
- work through GitHub issues labeled ai-task automatically
- run multiple AI coding agents in parallel on separate tasks
- enforce project coding rules on AI-generated code
- auto-create pull requests from AI agent work

## When to choose
- you want autonomous, unattended AI coding sessions driven by a task list or PRD
- you use multiple AI coding CLIs (Claude Code, Codex, Cursor, etc.) and want one orchestrator
- you need per-task git branches, auto-commits, and PR creation
- you want parallel agent execution isolated with git worktrees

## When to avoid
- you need a GUI or IDE-integrated AI assistant rather than a CLI loop
- you want fine-grained human review at every step rather than autonomous completion
- your project has no test or lint setup to validate agent output
- you require a permissive license - the repo has none

## Facets
- artifact type: cli-tool
- maturity: active
- function: agent-framework, cli, workflow-automation, developer-tools, ci-cd
- domain: developer-tools, large-language-models
- platform: cli, windows
- tags: ai-coding-agents, autonomous-loop, ralph-loop, claude-code, git-worktrees, prd-driven-development, code-generation, automation, ai-agents, command-line, nodejs, macos, linux

## Member repositories
- michaelshimeles/ralphy (main) score 45

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:32.445529+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T07:32:07.134898+00:00, confidence not recorded.
  - readme: https://github.com/michaelshimeles/ralphy (fetched 2026-08-28T04:07:32.445529+00:00, sha 98129902df8a)
  - homepage: https://ralphy.goshen.fyi (fetched 2026-08-29T09:47:34.434746+00:00, sha e7dee5b313ca)
- Data as of 2026-08-30T08:39:29.467469+00:00.