building-incident-response-playbook

Files (4)

• LICENSE
• SKILL.md
• api-reference.md · references/
• agent.py · scripts/

SKILL.md (first lines)

# Building Incident Response Playbooks

## When to Use

- Establishing or maturing an incident response program from scratch
- Documenting procedures for a new incident type after a novel attack
- Automating response workflows in a SOAR platform (Cortex XSOAR, Splunk SOAR)
- Preparing for compliance audits requiring documented IR procedures (SOC 2, PCI-DSS, HIPAA)
- Conducting a gap analysis of existing IR capabilities against specific threat scenarios

**Do not use** for one-time ad hoc investigations; playbooks are reusable procedure documents, not case-specific reports.

## Prerequisites

- Organizational risk assessment identifying top incident scenarios by likelihood and impact
- NIST SP 800-61r3 or SANS PICERL framework adopted as the organizational IR standard
- Asset inventory with business criticality ratings and data classification
- RACI chart defining roles: Incident Commander, SOC analysts, system administrators, legal, communications
- Existing detection capabilities inventory (SIEM rules, EDR detections, IDS signatures)
- SOAR platform access if building automated playbooks

## Workflow

### Step 1: Select and Scope the Incident Type

Define the specific scenario the playbook will address:

- Identify the top incident types based on organizational risk assessment and historical data
- Scope each playbook to a single incident type for clarity (do not combine unrelated scenarios)
- Define trigger conditions that activate the playbook

Common playbook types:
```
Priority Playbooks (build first):
1. Ransomware incident response
2. Phishing/credential compromise
3. Business email compromise
4. Malware infection
5. Data breach/exfiltration
6. DDoS attack
7. Insider threat
8. Account takeover
9. Web application compromise
10. Cloud infrastructure compromise
```

### Step 2: Define the Playbook Structure

Every playbook should follow a consistent structure:

```
PLAYBOOK TEMPLATE
━━━━━━━━━━━━━━━━
1. Playbook Metadata
   - Name, version, owner, last review date
   - Trigger conditions
   - Severity criteria

2. RACI Matrix
   - Who is Responsible, Accountable, Consulted, Informed for each step

3. Detection & Triage
   - How the incident is detected
   - Initial triage checklist
   - Severity classification criteria

4. Containment
   - Short-term containment actions
   - Long-term containment actions
   - Evidence preservation requirements

5. Eradication
   - Root cause identification
   - Malware/threat removal steps
   - Verification procedures

6. Recovery
   - System restoration steps
   - Validation criteria
   - Monitoring requirements post-recovery

7. Post-Incident
   - Lessons learned meeting trigger
   - Report template
   - Detection improvement actions

8. Communication
   - Internal notification matrix
   - External notification requirements (regulators, customers, law enforcement)
   - Status update cadence

9. Appendices
   - Tool-specific procedures
   - Contact lists
   - Evidence collection checklists
```

### Step 3: Write Decision Trees and Escalation Criteria

Define clear decision points with binary outcomes:

```
Detection Alert Received
├── Is the alert a true positive?
│   ├── YES → Classify severity
│   │   ├── P1 (Critical) → Page incident commander, begin containment immediately
│   │   ├── P2 (High) → Notify IR lead, begin investigation within 30 min
│   │   ├── P3 (Medium) → Queue for investigation within 4 hours
│   │   └── P4 (Low) → Document and investigate within 24 hours
│   └── NO → Document as false positive, tune detection rule
└── Cannot determine → Escalate to Tier 2 for deeper analysis
```

Escalation triggers:
- Any P1 incident: Immediate escalation to IR lead and CISO
- Data exfiltration confirmed: Legal counsel and privacy officer notified
- Customer data involved: Customer notification process activated
- Third-party involvement: Vendor security contact engaged
- Law enforcement needed: General counsel authorizes before contact

From adriannoes/awesome-agentic-ai, Apache-2.0.

More from adriannoes/awesome-agentic-ai