Agent Skills
› yaojingang/yao-meta-skill
› incident-command-governor
incident-command-governor
GitHub构建标准化事故指挥包,用于事故复盘、严重性评估及沟通组装。通过收集输入、参照政策文件,生成包含摘要、严重性、时间线、影响范围、负责人及沟通计划的治理型文档,确保流程可审计且信息完整。
Trigger Scenarios
需要标准化事故复盘时
执行严重性评估时
组装事故沟通内容时
Install
npx skills add yaojingang/yao-meta-skill --skill incident-command-governor -g -y
SKILL.md
Frontmatter
{
"name": "incident-command-governor",
"description": "Build governed incident command packets. Use when asked to standardize incident review, run severity assessment, or assemble incident communication."
}
Incident Command Governor
When To Use
- You need a reusable incident command packet from messy operational inputs.
- You need explicit severity assessment, stakeholder updates, and action ownership.
- You need a governed workflow that can be reviewed, maintained, and audited over time.
Do Not Use
- The task is only to explain an incident concept.
- The task is only to draft a one-off update.
- The request is still brainstorming and not yet packaging a repeatable incident workflow.
Workflow
- Gather the minimum incident inputs:
- incident summary
- timeline evidence
- affected systems and customers
- severity indicators
- owners and responders
- stakeholder audiences
- Read
references/severity-matrix.mdto assign or confirm severity. - Read
references/comms-policy.mdto shape internal and external updates. - Read
references/review-policy.mdbefore finalizing, because this skill is maintained under a monthly review cadence. - Use
scripts/build_incident_packet.pywithinput/incident_input_example.jsonwhen deterministic packet assembly is needed. - Produce one governed incident packet with:
- incident summary
- severity block
- evidence-backed timeline
- affected scope
- owner and action table
- internal and external comms blocks
- explicit open risks and review notes
- Stop and report blockers if severity evidence or ownership data is missing.
Output Contract
The final answer must include:
- a concise incident summary
- explicit severity and rationale
- a timeline section
- impact scope
- owners and next actions
- stakeholder communication blocks
- governance note for follow-up review
Validation Checklist
- Severity is justified with evidence, not guessed.
- Owner and next actions are explicit.
- Communications are separated by audience.
- Missing evidence is surfaced as a blocker.
- The packet can be reviewed again without reconstructing context from chat history.
Reference Map
- Read
references/severity-matrix.mdfor severity rules. - Read
references/comms-policy.mdfor communication guidance. - Read
references/review-policy.mdfor review cadence and governed lifecycle expectations. - Inspect
evals/trigger_cases.jsonfor routing boundaries. - Inspect
outputs/incident_packet_example.mdfor an example final artifact. - Inspect
reports/governance_score.jsonfor the current maturity snapshot.
Version History
- c37915e Current 2026-07-05 12:03


