Autonomous Loops

This is the practical guide to running Claude Code without babysitting it. It walks through the full spectrum from simple bash scripts chaining `claude -p` calls (perfect for daily dev workflows) to NanoClaw's persistent REPL sessions, all the way up to RFC-driven DAG orchestration with parallel agents and merge queues. The sequential pipeline pattern is especially clarifying: if you can't script it with `claude -p`, you probably can't drive it interactively either. Notable for actually showing working examples instead of theory, and the De-Sloppify pattern is a clever answer to Claude's tendency to over-test language features. Note the compatibility warning: this skill is being sunset in favor of continuous-agent-loop, so check there first for newer patterns.

skills.sh installs: 5.6K
GitHub stars: 300.3K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add affaan-m/ecc --skill autonomous-loops --agent claude-code

Installs into .claude/skills of the current project.

Files

Autonomous Loops Skill

Compatibility note (v1.8.0): autonomous-loops is retained for one release. The canonical skill name is now continuous-agent-loop. New loop guidance should be authored there, while this skill remains available to avoid breaking existing workflows.

Patterns, architectures, and reference implementations for running Claude Code autonomously in loops. Covers everything from simple claude -p pipelines to full RFC-driven multi-agent DAG orchestration.