Paste the prompt.
Your agent fetches the Loopify skill and follows it.
LOOPMASTER
Agents work in loops.
They repeat a cycle: look at the job, decide the next step, act, check the result, until a stop condition is met or there is no useful work left. Loopmaster teaches you how agent loops work, and gives your agent a skill that builds one for you.
Fetch https://loopmaster-ai.pages.dev/skills/loopify/SKILL.md and follow it.
Interview me, one question at a time, to define a loop for a job I want
done: the goal, the steps to repeat, how to check each pass worked, and
when to stop. Don't set up or schedule anything until I approve the plan.Works with Claude, Codex, Hermes, or any agent that can fetch a URL.
01 / THE LOOP
What's an agent loop?
A loop is the oldest idea in programming: repeat until a condition says stop.
An agent loop is what happens when the thing repeating is an agent. Each pass, it looks at the state of the job, decides the next step, does it with a tool, and checks the result. Then it goes around again until the goal is met, or it cannot find any more useful work to do.
The model call is one part. The loop is the working system.
02 / LOOPIFY
Loopify: describe a job, get a loop.
Answer its questions.
One at a time: what is the job, what are the steps, what proves a pass worked, when should it stop, what needs your approval.
Get your loop.
A clear written plan with steps, checks, a stop condition, and human gates. Your agent can run it only after you approve it.
03 / LEARN
Learn the loop.
What is an agent loop?The 30-second and five-minute answer.5 minHow loops endStop conditions, step limits, and why stopping is the hard part.4 minLoops in the wildCoding agents, research agents, and monitors share the same shape.4 minBuild your own loopDefine one by hand, or let Loopify interview you.3 min
04 / READING
What the field is saying.
The best writing on agent loops by the people who build them. Hand-picked, one line each on why it is worth your time.
Anthropic: Building Effective AgentsThe canonical workflow-versus-agent framing, with the simple-loop-first advice every builder should start from.↗Yao et al.: ReAct: Synergizing Reasoning and Acting in Language ModelsThe paper that made the think, act, observe cycle legible as the core shape of modern agents.↗Addy Osmani: Loop EngineeringA practitioner-friendly explanation of designing the recurring system around the model, not just prompts.↗Simon Willison: Designing agentic loopsA short, sharp guide to tool choice and stop conditions from one of the clearest writers in the field.↗Simon Willison: How coding agents workThe cleanest tour of the loop inside coding agents like Claude Code and Codex.↗Sydney Runkle (LangChain): The Art of Loop EngineeringShows how agent, verification, event, and improvement loops stack around each other in real systems.↗
05 / SKILLS