looper
Design safe, reviewable AI agent loops before running them
All Skills in This Repository (2)
Pure Emerald Level Indicatorslooper
Designs agent loops with explicit goals, verification criteria, and termination conditions for Claude Code users.
looper
Designs and scaffolds agent loops with verification gates, cross-model reviewers, and termination guards.
Frequently Asked Questions
FAQPage SchemaHow to install Looper?โผ
Run `npx skills add ksimback/looper --all -g -y` in your terminal to install the skill globally for your AI agent.
What does Looper do?โผ
Looper interviews you to design a solid agent loop with a clear goal, verification checks, and a reviewer model, then generates all the files needed to run it.
How is Looper different from /goal or /loop?โผ
Those commands only run loops; Looper is the design layer in front of them that critiques your goal, adds typed verification, and wires in a second model as judge before anything runs.
Can Looper use non-Claude models as reviewers?โผ
Yes. It detects installed model CLIs and lets you assign a different model family as reviewer or judge, with consent and redaction controls for any data leaving your machine.
Do I need coding skills to use Looper?โผ
No. You answer a guided interview in plain language, and Looper writes the loop spec, runner script, and execution prompt for you.
Related Repositories in Software Engineering
View All in Software Engineeringโopenclaw
Run a personal AI assistant across your devices and chat apps
superpowers
Gives coding agents a disciplined workflow from idea to merged code
react
AI agent skills for building, testing, and porting React core