ralph

Orchestrate repeated agent iterations to complete in-progress feature plans.

17|2|Updated Jan 31, 2026

Search Tags:#agents #iteration #autonomy #task-management #plan-execution #progress-tracking

Ready to run

npx skills add https://github.com/colmarius/dot-agents --skill ralph-colmarius

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: ralph
Source: https://github.com/colmarius/dot-agents/tree/main/.agents/skills/ralph
Command: npx skills add https://github.com/colmarius/dot-agents --skill ralph-colmarius

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Ralph automates multi-iteration feature development by coordinating repeated agent work cycles so that tasks progress toward completion with minimal human intervention.

Core Features & Use Cases

Orchestrates autonomous code generation through repeated agent instances until all tasks in a plan are complete.
Tracks progress with in-plan status markers and a companion progress file, pausing for review when needed.
Enforces disciplined development practices: per-task commits, explicit footers, and safe scope boundaries.

Quick Start

Run the ralph skill on a plan file such as .agents/plans/in-progress/my-feature.md to begin autonomous, iterative execution.

Frequently Asked Questions about ralph

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

FAQPage Schema

How do I automate autonomous feature development across multiple agent iterations?▼

Autonomous feature development is automated by orchestrating repeated agent iterations until all tasks in a plan are complete. The execution loop parses plan files, tracks per-task status, enforces commit discipline, and pauses for review when needed to maintain control.

What is the best way to track progress for iterative task execution in software projects?▼

Progress tracking for iterative task execution is handled through in-plan status markers and a companion progress file. This mechanism monitors per-task completion states and triggers pause-for-review workflows to ensure safe handoffs and auditability.

How does plan parsing work for autonomous agent task management?▼

Plan parsing for autonomous task management extracts in-progress task lists from plan files to drive the execution loop. It evaluates stop conditions, assigns per-task statuses, and enforces commit discipline with explicit footers to maintain safe scope boundaries.

Can I pause autonomous code generation for review during plan execution?▼

Yes, autonomous code generation can pause for review during plan execution. The workflow monitors progress against stop conditions and safe handoff boundaries, allowing you to intervene, audit changes, and resume iterative development without losing track.

Do I need a specific plan file format to start autonomous iterative development?▼

You need a plan file with in-progress task lists, such as a markdown file in a plans directory. The execution loop parses this plan to extract tasks, apply per-task status markers, and enforce commit discipline throughout the autonomous development cycle.

What are the limitations of autonomous iterative execution for software projects?▼

Autonomous iterative execution requires explicit stop conditions and per-task commit discipline to prevent scope creep. It is limited by the plan file's structure and pauses for review when safe handoffs are needed, meaning poorly defined tasks will halt progress.