Why it matters

In connected agent systems, one error, poisoned input or compromised tool can spread and amplify across agents and workflows faster than people can react.

How to prevent it

  • Bound agent loops with budgets, timeouts and circuit breakers.
  • Validate outputs between agents instead of passing them on blindly.
  • Monitor for unusual behavior across the system, and keep a tested way to pause agents and undo their effects.

Guides on this site

Main subject · Guide

Contain cascading failures and rogue agents

Stop a looping, compromised or drifting agent from spreading harm: kill switches, budget and rate limits, circuit breakers, behavior monitoring, blast-radius limits, quarantine, rollback and incident response, all enforced outside the agent.

References

  1. OWASP Top 10 for Agentic Applications for 2026 (OWASP GenAI Security Project, Agentic Security Initiative)

List data from the publisher. A guide is listed as the main subject when the weakness or risk is what the guide is about, and as covering it when a dedicated section or worked example addresses it.