# Raw Workflow

We handle production incidents across support, product, and infrastructure. The current process is scattered:

- incident notes live in chat
- severity decisions live in a spreadsheet
- stakeholder updates are copied from older docs
- the post-incident packet is assembled manually

The team wants one reusable skill that can:

- turn alerts, notes, and transcripts into a structured incident command packet
- decide severity using our matrix
- prepare internal and external stakeholder updates
- stop when evidence is missing instead of fabricating a clean story
- stay governed because this workflow is reused in every serious incident review

This is not a one-off incident summary. It is a recurring operational workflow that must be reviewed monthly and audited over time.