DevOps自动化运维Agent任务编排提示词

You are a senior DevOps automation architect. Help me design an autonomous DevOps agent workflow for the following infrastructure scenario.

Context

I need to create an automated agent that can:

1. Monitor application health and infrastructure metrics
2. Detect anomalies and potential failures
3. Execute predefined remediation playbooks
4. Scale resources based on demand
5. Report actions taken and escalate when needed

Please Design:

Task Graph

Define the agent task flow as a DAG (Directed Acyclic Graph):

• Trigger conditions for each task
• Dependencies between tasks
• Timeout and retry policies
• Rollback procedures

Monitoring Rules

Define alert conditions, thresholds, and response actions in YAML format.

Remediation Playbooks

For each common failure scenario:

• Detection criteria
• Step-by-step automated response
• Human escalation triggers
• Success/failure validation

Scaling Strategy

• Metrics-based auto-scaling rules
• Predictive scaling based on historical patterns
• Cost optimization constraints
• Graceful scale-down procedures

Safety Guardrails

• Maximum blast radius per automated action
• Required approvals for destructive operations
• Audit logging requirements
• Circuit breaker conditions

My Infrastructure:

[Describe your stack: cloud provider, orchestration platform, key services, current pain points]

Please generate a complete agent task orchestration plan with YAML/JSON configs where appropriate.