Files
yao-meta-skill/examples/governed-incident-command/generated-skill/SKILL.md
T
2026-03-31 23:00:49 +08:00

71 lines
2.6 KiB
Markdown

---
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
1. Gather the minimum incident inputs:
- incident summary
- timeline evidence
- affected systems and customers
- severity indicators
- owners and responders
- stakeholder audiences
2. Read `references/severity-matrix.md` to assign or confirm severity.
3. Read `references/comms-policy.md` to shape internal and external updates.
4. Read `references/review-policy.md` before finalizing, because this skill is maintained under a monthly review cadence.
5. Use `scripts/build_incident_packet.py` with `input/incident_input_example.json` when deterministic packet assembly is needed.
6. 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
7. 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.md` for severity rules.
- Read `references/comms-policy.md` for communication guidance.
- Read `references/review-policy.md` for review cadence and governed lifecycle expectations.
- Inspect `evals/trigger_cases.json` for routing boundaries.
- Inspect `outputs/incident_packet_example.md` for an example final artifact.
- Inspect `reports/governance_score.json` for the current maturity snapshot.