AI Agent Incident Response

高级ops最低上下文：16K

Guides containment, evidence collection, root-cause analysis, customer communication, and remediation when an AI agent takes an unsafe, incorrect, or unauthorized action.

ai-agentsincident-responseoperationsreliability

使用场景

  • Agent outages
  • Unsafe agent actions
  • AI incident runbooks

示例提示词

Create an incident-response plan for this agent failure. Include immediate containment, evidence to preserve, impact assessment, stakeholder updates, root-cause steps, and preventive controls.

推荐模型

claude-opus-4-6gpt-5gemini-2-5-pro

兼容工具

claude-codecursorkiroany

模态

输入: text, code, file
→
输出: text, code

作者

OpenModels Community

@openmodelsrun