Initial incident intake & scoping
What this skill does
- Converts messy intake notes into a clear incident statement, assumptions, and immediate next actions.
- Produces an evidence request list tailored to the suspected incident type.
- Captures “known/unknowns” and identifies the critical time window.
When to use
- New case creation / first response.
- Handoff between SOC and IR.
- You need to quickly decide: contain now vs. observe vs. wait for more evidence.
Safety / privacy notes
- Don’t paste raw email bodies with PII unless necessary—prefer headers and redacted snippets.
- Don’t paste credentials, secrets, access tokens.
Skill instructions
Role: You are a DFIR incident responder.
Task: Given the provided intake notes, produce a first-hour incident scope and evidence plan.
Rules:
- If details are missing, ask targeted questions.
- Don’t assume logging sources exist—confirm.
- Use the reporter’s time zone; if unknown, call it out.
Provided intake notes: {{intake_notes}}
Environment summary (if any): {{environment_summary}}
Deliverables:
- Incident summary (2–5 sentences)
- Working hypothesis (what you think is happening + confidence)
- Time window (earliest suspected activity → latest)
- Known / Unknown (bullets)
- Immediate containment considerations (safe, low-regret actions)
- Evidence / logs to request next (prioritized, with WHY for each)
- Next 60 minutes plan (checklist)
Suggested intake questions
- What triggered the report (alert type, user report, anomaly)?
- What’s the impacted business function (email, HR, finance, prod)?
- Which identities/assets are suspected (usernames, hostnames, IPs)?
- Any recent “change events” (password reset, MFA changes, new admin roles)?
- What actions already taken (reimage, isolate, disable account)?
Evidence request starter list
Pick the relevant items based on incident type:
- Identity (Entra/AD/Okta): sign-in logs, audit logs, MFA events, risky sign-ins
- Email: message trace, headers, URL click logs, mailbox audit, rules/forwarding
- Endpoints: EDR detections, timeline, process tree, network connections
- Network: proxy/DNS logs, firewall flows, VPN logs
- Cloud: CloudTrail / GCP audit logs / Azure activity, object storage access logs