R

Ralph

Autonomous loop that runs Claude Code until your whole feature list is done

PriceFree / Open SourceReviewed2026-04-12

Expert verdict

Ship

2-1
2 Ships1 Skips
Visit github.com

The Panel's Take

Ralph is an open-source TypeScript tool that runs AI coding agents (Claude Code or Amp) in repeated cycles until every story in a Product Requirements Document is complete. Each iteration gets a fresh context window, but Ralph maintains institutional memory through git commits, a progress.txt file tracking learnings, and a prd.json tracking task status. It runs quality gates (typecheck + tests) before marking a story done and looping to the next. 15.8k stars and currently trending — it's a viral implementation of Geoffrey Huntley's 'Ralph pattern' for autonomous multi-story development.

Share this verdict

Ralph verdict: SHIP 🚀

2 ships · 1 skip from the expert panel

Full review: shiporskip.io/tool/ralph-autonomous-ai-agent-loop-prd-claude-code-amp-cycles-2026

Weekly AI Tool Verdicts

Get the next verdict in your inbox

7 critics review a new AI tool every day. Weekly digest — free.

Looking for Ralph alternatives?

Compare Ralph with every other Developer Tools tool reviewed by our panel.

See all Developer Tools alternatives

Embed this verdict

Tool makers can add a live ShipOrSkip badge to their site. Badge loads track impressions; clicks route back to this review.

Ship · 6.7/10
HTML badge
<a href="https://shiporskip.io/api/badge-click/ralph-autonomous-ai-agent-loop-prd-claude-code-amp-cycles-2026" target="_blank" rel="noopener"><img src="https://shiporskip.io/api/badge/ralph-autonomous-ai-agent-loop-prd-claude-code-amp-cycles-2026" alt="Ralph Ship verdict on ShipOrSkip" width="360" height="90" /></a>
Markdown badge
[![Ralph Ship verdict on ShipOrSkip](https://shiporskip.io/api/badge/ralph-autonomous-ai-agent-loop-prd-claude-code-amp-cycles-2026)](https://shiporskip.io/api/badge-click/ralph-autonomous-ai-agent-loop-prd-claude-code-amp-cycles-2026)
Iframe widget
<iframe src="https://shiporskip.io/embed/ralph-autonomous-ai-agent-loop-prd-claude-code-amp-cycles-2026" title="Ralph ShipOrSkip verdict" width="360" height="260" style="border:0;border-radius:16px;max-width:100%;" loading="lazy"></iframe>

The reviews

The fresh-context-per-cycle approach solves the single biggest problem with AI coding agents: context exhaustion on multi-hour tasks. The prd.json format enforces the right discipline — stories small enough for one context window, outcomes defined in advance. I've shipped three features with this and it works as advertised when you write good PRDs.

Helpful?

Ralph's fatal flaw is that it's only as good as your PRD, and writing a perfect PRD is harder than just coding the feature yourself. The quality gates catch compile errors but not logic bugs — you can come back to 20 commits of plausible-looking garbage that all passes typecheck. This works on toy projects, not production codebases.

Helpful?

15.8k stars in what appears to be weeks is a signal that the market was waiting for exactly this — a simple, composable loop over AI agents. Ralph isn't the final form, but the pattern is the future. Expect Cursor, Windsurf, and Claude Code itself to absorb this workflow natively within the year.

Helpful?

For non-devs who can write a PRD but not code, Ralph is genuinely unlocking: describe what you want, let it run overnight, review the PR. The CLI UX is minimal but that's fine. The real experience is in the progress.txt file, which is weirdly satisfying to read — like watching an AI developer take notes.

Helpful?

Bookmarks

Loading bookmarks...

No bookmarks yet

Bookmark tools to save them for later