skills.rest
trevoraspencer
trevoraspencerCOMMUNITY
Explore all skills by this author →
GitHub

ralph

Orchestrate autonomous coding loops from feature specs with persistent state.

Updated Aug 23, 2026
Search Tags:#automation#workflow#agent#coding#feature-implementation#loop#ralph
One-click install
npx skills add https://github.com/trevoraspencer/skill-ralph-loop-creator --skill ralph-trevoraspencer
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ralph
Source: https://github.com/trevoraspencer/skill-ralph-loop-creator/tree/main
Command: npx skills add https://github.com/trevoraspencer/skill-ralph-loop-creator --skill ralph-trevoraspencer

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires jq, git, and includes scripts (resource) components.

What problem does it solve?

Ralph enables teams to automate the creation of autonomous coding loops that break features into atomic user stories and execute them iteratively with fresh agent contexts.

Core Features & Use Cases

  • Automates reading a feature spec, asking clarifying questions, and selecting execution agents and models.
  • Generates a prd.json with prioritized user stories and a loop script under .ralph/<loop-name>.sh, then runs one story per iteration with state persisted across runs.
  • Supports forward mode for end-to-end implementation and decompose mode for breaking existing features into tasks.

Quick Start

Run Ralph in an interactive session with /ralph @plan.md to start executing feature stories one at a time.

Frequently Asked Questions about ralph

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate feature implementation with autonomous coding agents?▼

Automate feature implementation by orchestrating autonomous agent loops that read feature specs, decompose requirements into atomic user stories, and execute them iteratively with fresh agent contexts and persistent state.

What is the best way to break feature specs into executable coding loops?▼

Breaking feature specs into executable coding loops is done by generating a prioritized prd.json and a loop script, then running one story per iteration to guide agent choice and per-iteration execution.

Do I need git and jq to run autonomous agent workflows?▼

Git and jq are required dependencies to run autonomous agent workflows, supporting state persistence and script execution for orchestrating feature-driven coding loops.

Can I automate pull request creation during feature loop execution?▼

Optional PR automation is supported during feature loop execution, allowing autonomous agents to manage end-to-end workflow orchestration and generate pull requests upon completing iterations.

How does decompose mode differ from forward mode in feature workflow orchestration?▼

Decompose mode breaks existing features into tasks, while forward mode runs end-to-end implementation, both utilizing persistent state and fresh headless agent contexts for per-iteration execution.

How do I start an interactive session to execute feature stories one at a time?▼

Start an interactive session by invoking the agent with a feature plan file, which triggers the generation of user stories and begins executing feature stories one iteration at a time.

Related Skills

Explore More Software Engineering Skills→
misadecastromisadecastro

speckit-specify

Generate feature specifications and checklists from natural language descriptions.

Community
Advanced
Lucas23-IECILucas23-IECI
1

ralph

Automate feature implementation by executing user stories in isolated git-tracked iterations.

Community
Advanced
praveenpererapraveenperera
4

review-fix-loop

Automate iterative GitHub pull request review cycles until no issues remain.

Community
Advanced
automatedigitalautomatedigital
1

opencode

Delegate coding tasks to the OpenCode CLI autonomous agent.

Official
Intermediate