CyclesRuntime authority for autonomous agents

Keep agents within approved spend, risk, and action limits. Multi-language, Open protocol, Apache 2.0.

What is Cycles?

End-to-End Tutorial

View Demos

Add runtime authority in a few lines

@cycles reserves budget before the action runs. No remaining cycles — no action.

from runcycles import cycles

@cycles(estimate=5000, action_kind="llm.completion", action_name="openai:gpt-4o")
def ask(prompt: str) -> str:
    return openai.chat.completions.create(
        model="gpt-4o",
        messages=[{"role": "user", "content": prompt}]
    ).choices[0].message.content

Quickstart Guides

Get started in Python, TypeScript, Spring Boot, or via MCP for Claude / Cursor / Windsurf. Deploy the full stack with Docker in under 5 minutes.

Get started

Concepts

Understand why rate limits fail for agents, what exposure means for autonomous systems, and how teams evolve from dashboards to runtime authority.

Learn more

How-To Guides

Integrate with OpenAI, Anthropic, LangChain, Vercel AI SDK, AWS Bedrock, and more. Budget patterns, estimation strategies, error handling, and testing.

Browse guides

Protocol Reference

Reserve-commit lifecycle, authentication, scope derivation, units, caps, overage policies, TTL, decide, dry run, events, debt, balances, metrics, and error handling.

Read reference

Incident Patterns

Real failure modes — runaway agents, tool loops, retry storms, concurrent overspend, and scope misconfigurations.

See patterns

Operations

Production deployment, monitoring, security hardening, and troubleshooting guides for platform teams.

View guides