index.md

layout	home
title	Cycles — Runtime authority for autonomous agents
titleTemplate	false
description	Runtime authority for autonomous AI agents — pre-execution gating on cost, risk, and tool actions. Open protocol, self-hosted, Apache 2.0.
hero	name text tagline image actions Cycles Runtime authority for autonomous agents Stop runaway agent spend and risky actions before they execute.<br><span style="font-size: 0.85em; opacity: 0.7;">Self-hosted, no prompt storage. Apache 2.0.</span> light dark alt /runcycles-hero.svg?v=7 /runcycles-hero-dark.svg?v=7 Cycles hero — agents → gate → allow/block → signed audit ledger theme text link brand Start with Claude or Codex /how-to/add-cycles-with-claude-or-codex theme text link alt Run the demos /demos/ theme text link alt Evaluate your stack /how-to/evaluate-cycles-for-agent-saas
features	icon title details link linkText src alt /icons/rocket.svg Quickstart icon Quickstart Guides Get started in Python, TypeScript, Rust, Spring Boot, or via MCP for Claude / Cursor / Windsurf. Deploy the full stack with Docker in under 5 minutes. /quickstart/what-is-cycles Get started icon title details link linkText src alt /icons/info-circle.svg Concepts icon Concepts Understand why rate limits fail for agents, what exposure means for autonomous systems, and how teams evolve from dashboards to runtime authority. /concepts/why-rate-limits-are-not-enough-for-autonomous-systems Learn more icon title details link linkText src alt /icons/book-text.svg Topic guides icon Topic Guides Two reference maps covering the enforcement dimensions of runtime authority: LLM cost control (what agents spend) and risk & blast radius (what agents do). /guides/ Read the guides icon title details link linkText src alt /icons/gauge.svg Tools icon Tools Interactive tools that quantify the case for runtime authority — a Claude-vs-GPT cost calculator and an AI-agent blast-radius risk calculator. Shareable URLs, embeddable, branded export. /calculators/ Open tools icon title details link linkText src alt /icons/wrench.svg How-To icon How-To Guides Add Cycles with Claude or Codex from a single prompt, or use language- and framework-specific guides for Python, TypeScript, Spring, MCP, LangChain, LangGraph, OpenAI Agents, and more. Budget patterns, estimation, error handling. /how-to/add-cycles-with-claude-or-codex Browse guides icon title details link linkText src alt /icons/book-text.svg Protocol icon Protocol Reference Reserve-commit lifecycle, authentication, scope derivation, units, caps, overage policies, TTL, decide, dry run, events, debt, balances, metrics, and error handling. /protocol/how-reserve-commit-works-in-cycles Read reference icon title details link linkText src alt /icons/alert-triangle.svg Incidents icon Incident Patterns Real failure modes — runaway agents, tool loops, retry storms, concurrent overspend, and scope misconfigurations. /incidents/runaway-agents-tool-loops-and-budget-overruns-the-incidents-cycles-is-designed-to-prevent See patterns icon title details link linkText src alt /icons/settings.svg Operations icon Operations Production deployment, monitoring, security hardening, and troubleshooting guides for platform teams. /how-to/production-operations-guide View guides