autonomous-loops

Automate Claude Code loop design across sequential, REPL, and DAG patterns.

2|Updated Mar 20, 2026
One-click install
npx skills add https://github.com/multiplex-ai/muggle-ai-teams --skill autonomous-loops-multiplex-ai
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: autonomous-loops
Source: https://github.com/multiplex-ai/muggle-ai-teams/tree/main/skills/autonomous-loops
Command: npx skills add https://github.com/multiplex-ai/muggle-ai-teams --skill autonomous-loops-multiplex-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Patterns and architectures for autonomous Claude Code loops — from simple sequential pipelines to RFC-driven multi-agent DAG systems.

Core Features & Use Cases

  • Sequential Pipeline (claude -p) for simple, repeatable steps.
  • NanoClaw REPL for persistent, session-based Claude Code exploration.
  • Infinite Agentic Loop for spec-driven parallel generation with waves.
  • Continuous Claude PR Loop for automated iteration, CI checks, and landings.
  • De-Sloppify Pattern for post-implement cleanup passes.
  • Ralphinho / RFC-Driven DAG Orchestration for large, multi-unit RFCs with merge queues.

Quick Start

Start with the Sequential Pipeline example to pilot a basic loop using claude -p and iterate from there.

Frequently Asked Questions about autonomous-loops

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

FAQPage Schema
What is an autonomous Claude Code loop and how does it work?

An autonomous Claude Code loop minimizes human intervention by automating sequential pipelines, REPL sessions, and RFC-driven DAG orchestrations with context persistence and quality gates.

How do I build a sequential pipeline with Claude Code for repeatable steps?

Start with the Sequential Pipeline pattern using claude -p to automate simple, repeatable steps. Iterate from this baseline to establish your autonomous Claude Code loop.

Can I orchestrate multi-agent workflows using RFC-driven DAGs in Claude Code?

Yes, the RFC-Driven DAG Orchestration pattern supports large, multi-unit RFCs by orchestrating multi-agent workflows with isolated stages, explicit merge strategies, and a merge queue.

What's the best way to automate CI checks and landings for Claude Code pipelines?

Use the Continuous Claude PR Loop pattern to automate iteration, run CI checks, and manage landings through a merge queue strategy across isolated pipeline stages.

Does Claude Code support persistent session-based exploration for autonomous workflows?

Yes, the NanoClaw REPL pattern provides persistent, session-based Claude Code exploration to maintain context across autonomous workflow stages.

How do I clean up generated code after running parallel agentic loops?

Apply the De-Sloppify Pattern to execute post-implement cleanup passes after the Infinite Agentic Loop completes its spec-driven parallel generation waves.