AgentSkillsCN

crisis-manager

在生产故障或重大安全事件发生时，快速响应并协调诊断、临时修复以及事后数据收集工作。

SKILL.md
--- frontmatter
name: crisis-manager
description: Provides rapid response during production incidents or critical security breaches. Coordinates diagnostics, temporary fixes, and post-mortem data collection.
status: implemented
arguments:
  - name: dir
    short: d
    type: string
    default: "."
    description: Project/repository directory
  - name: log
    short: l
    type: string
    description: Path to log file to analyze
  - name: since
    short: s
    type: string
    default: 24 hours ago
    description: Analyze changes since this time
  - name: out
    short: o
    type: string
    description: Output file path

Crisis Manager

This skill is for high-stakes, time-sensitive situations where rapid recovery is paramount.

Capabilities

1. Incident Diagnostic

  • •Rapidly correlates logs, security alerts, and recent commits to find the "smoking gun."
  • •Utilizes the "Three Pillars of Observability" (Metrics, Logs, Traces) as defined in Monitoring Best Practices.
  • •Suggests immediate workarounds or rollbacks.

2. Post-Mortem Preparation

  • •Captures the state of the system during the incident for later analysis.
  • •Drafts the initial incident report (What happened, Timeline, Immediate Action) following PagerDuty Best Practices for blameless post-mortems.

Usage

  • •"We have a production outage! Run crisis-manager to analyze logs and recent changes immediately."
  • •"A critical zero-day was found. Coordinate with security-scanner to find all affected instances."

Knowledge Protocol

  • •This skill adheres to the knowledge/orchestration/knowledge-protocol.md. It automatically integrates Public, Confidential (Company/Client), and Personal knowledge tiers, prioritizing the most specific secrets while ensuring no leaks to public outputs.
  • •Integrates PagerDuty Best Practices for incident roles (e.g., Incident Commander) and resolution workflows.
  • •References Runbook Best Practices for executing machine-readable diagnostic and remediation steps.