Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
111 changes: 107 additions & 4 deletions skills/incident-response/ir-playbook/SKILL.md
Original file line number Diff line number Diff line change
Expand Up @@ -10,10 +10,10 @@ description: >
tags: [incident-response, ir, playbook]
role: [soc-analyst, security-engineer, vciso]
phase: [respond, recover]
frameworks: [NIST-SP-800-61r2, SANS-IH]
frameworks: [NIST-SP-800-61r2, SANS-IH, SEC-Item-1.05]
difficulty: intermediate
time_estimate: "30-60min"
version: "1.0.1"
version: "1.0.2"
author: unitoneai
license: MIT
allowed-tools: Read, Grep, Glob
Expand Down Expand Up @@ -61,6 +61,9 @@ Before beginning, gather or confirm the following. Mark each item as obtained or
- [ ] **Current state** -- Is the attack ongoing, contained, or resolved? What actions have already been taken?
- [ ] **Existing IR plan** -- Does the organization have a documented IR plan, designated IR team, and established communication channels?
- [ ] **Regulatory obligations** -- Applicable breach notification requirements (GDPR 72-hour rule, HIPAA, state breach notification laws, SEC 4-day rule, PCI DSS).
- [ ] **Materiality evidence** -- Functional, financial, operational, legal/regulatory, customer, and reputational impact data needed for a public-company materiality determination.
- [ ] **Cloud identity/session scope** -- SaaS sessions, OAuth grants, refresh tokens, service accounts, access keys, role sessions, and API credentials that could continue exfiltration after host isolation.
- [ ] **Deepfake / synthetic social engineering indicators** -- Voice, video, chat, or synthetic identity evidence; requested payment/access action; and out-of-band verification status.
- [ ] **Third-party dependencies** -- Managed security providers (MSSP/MDR), cyber insurance carrier notification requirements, external IR retainer.

---
Expand Down Expand Up @@ -127,6 +130,7 @@ Classify the incident using the NIST SP 800-61 taxonomy:
| **Supply Chain Compromise** | Compromise via trusted third-party software or service | Malicious update, compromised dependency, vendor breach |
| **Web Application Attack** | Exploitation of web application vulnerabilities | SQL injection, XSS, SSRF, API abuse |
| **Social Engineering** | Manipulation of personnel to gain access or information | Phishing, BEC, vishing, pretexting |
| **Synthetic Social Engineering** | AI-generated or manipulated voice, video, chat, or identity artifact used to induce payment, access, or trust decisions | Deepfake audio BEC, fake executive video call, synthetic vendor identity |

#### Step 2.2: Severity Determination

Expand Down Expand Up @@ -168,6 +172,32 @@ Assign severity based on the combination of functional impact, information impac
| **SEV-3 (Medium)** | Low functional impact OR information impact with regular recovery | IR team investigates during business hours; management notification within 24 hours |
| **SEV-4 (Low)** | None/minimal functional impact; no information impact; regular recovery | Documented and monitored; addressed in normal operations |

#### Step 2.2b: Materiality and Privilege Evidence Gate

For public-company, regulated, customer-impacting, or board-level incidents, gather a materiality package early. The SEC 4-business-day disclosure clock starts after materiality is determined, so the playbook should help counsel and executives make a documented determination rather than letting the issue drift.

```
IR-MAT-01: Potentially material incident lacks functional, financial, operational, legal/regulatory, customer, or reputational impact evidence
IR-MAT-02: SEC Item 1.05 clock owner, determination timestamp, or Form 8-K decision status is missing where public-company materiality may apply
IR-MAT-03: Internal executive notification lacks legal privilege / work-product handling guidance for SEV-1 or counsel-led incidents
IR-MAT-04: Cloud/SaaS exfiltration is routed only to network containment and omits token, session, OAuth grant, service-account, or API-key containment
IR-MAT-05: Wiper recovery proceeds without immutable-backup integrity, malware scan, restore-test, or backup age evidence
IR-MAT-06: Social engineering classification misses synthetic identity, deepfake audio/video, or out-of-band verification evidence
IR-MAT-07: Missing materiality, cloud-token, backup-integrity, deepfake, or privilege evidence is not marked Not Evaluable
```

**Materiality checklist:**

| Evidence Area | Questions | Decision Evidence |
|---|---|---|
| Functional impact | Are critical operations, production, revenue systems, or safety processes degraded? | Services affected, downtime, recovery estimate |
| Financial impact | Are revenue, costs, ransom/extortion, insurance, market, or contractual penalties material? | Finance estimate, loss range, assumptions |
| Legal/regulatory impact | Are SEC, GDPR, HIPAA, PCI, sector, or contractual notices triggered? | Counsel decision, notification deadline matrix |
| Customer/reputational impact | Are customers, partners, public services, or market confidence affected? | Customer count, communications plan, media risk |
| Governance decision | Who owns materiality, when was it evaluated, and what is the decision? | Timestamp, participants, decision, next review |

Mark the materiality decision **Not Evaluable** when impact evidence is incomplete or counsel/executive decision ownership is missing.

#### Step 2.3: Indicator Analysis

For each IOC, document and cross-reference:
Expand Down Expand Up @@ -198,6 +228,7 @@ START: Is the attack actively ongoing?
| |
| +-- YES --> IMMEDIATE CONTAINMENT
| | - Network isolation (disable switchport / security group)
| | - Revoke cloud/SaaS sessions, OAuth grants, API keys, and service-account credentials if valid tokens are involved
| | - Block egress to C2 IPs/domains at firewall
| | - Capture memory before power-off if possible
| | - Notify legal (potential breach notification trigger)
Expand Down Expand Up @@ -230,6 +261,8 @@ START: Is the attack actively ongoing?
- Rebuild from known-good baseline
```

**Cloud/SaaS exfiltration containment rule:** If the attacker is using valid cloud, SaaS, OAuth, API, or service-account credentials, network isolation of the endpoint is not sufficient. Containment must include token/session revocation, OAuth grant removal, service-account key or role-session review, downstream app checks, and sign-in/API log validation.

#### Step 3.1b: Wiper / Destructive Malware Response Track

Wiper malware destroys data irrecoverably (unlike ransomware which preserves encrypted data for ransom). This demands a fundamentally different response posture.
Expand All @@ -241,6 +274,16 @@ Wiper malware destroys data irrecoverably (unlike ransomware which preserves enc
3. **Verify backup integrity** -- Wipers target Volume Shadow Copies, backup agents, and NAS/SAN. Confirm offline/immutable backups exist before recovery planning.
4. **Preserve one affected system** (powered off, disk intact) for forensics and attribution.

**Cold shutdown and recovery integrity gate:**

| Gate | Required Evidence | Decision |
|---|---|---|
| Backup immutability | Offline/WORM/object-lock status, backup account separation, last successful immutable copy | Pass / Gap / Not Evaluable |
| Backup malware scan | AV/EDR/YARA scan of backup image or restored staging copy | Pass / Gap / Not Evaluable |
| Restore test | Isolated sandbox canary restore, hash/config comparison, boot/app smoke test, no production network access | Pass / Gap / Not Evaluable |
| Backup age and blast radius | Last known-good timestamp, systems covered, replication status, excluded assets | Pass / Gap / Not Evaluable |
| Re-wipe loop prevention | Persistence scan, scheduled task/GPO/startup review, backup-agent trust review | Pass / Gap / Not Evaluable |

**Key differences from ransomware:**

| Factor | Ransomware | Wiper / Destructive |
Expand Down Expand Up @@ -281,6 +324,7 @@ Use the appropriate communication template based on the audience.
**Internal Executive Notification (SEV-1/SEV-2):**

```
Privilege Handling: Attorney-Client Privilege / Attorney Work Product where directed by counsel
Subject: [SEVERITY] Security Incident - [Category] - [Incident ID]

Status: [Active | Contained | Eradicated | Recovered]
Expand All @@ -293,8 +337,11 @@ Data Impact: [Type and estimated volume of data affected, if applicable]
Current Actions: [What the IR team is doing now]
Next Update: [Scheduled time for next update]
Incident Commander: [Name and contact]
Distribution: [Need-to-know recipients; avoid compromised channels]
```

Privilege labels should be directed by counsel, tied to legal advice or work-product preparation, and paired with a named counsel owner and restricted distribution list. Do not over-label routine operational updates as privileged; track privileged and non-privileged communications separately.

**Legal/Regulatory Notification:**

```
Expand Down Expand Up @@ -343,6 +390,8 @@ Escalate to the next tier when any of the following conditions are met:
| Incident duration exceeds 4 hours without containment | IR lead escalates to management for resource allocation | At 4-hour mark |
| Evidence of supply chain compromise affecting customers | Legal, Customer communications, Executive leadership | Within 2 hours |
| Regulatory notification deadline approaching | Legal counsel, Compliance team | 24 hours before deadline |
| Potential SEC materiality determination needed | Legal counsel, CFO/Finance, Disclosure Committee, Executive leadership | Immediately; track 4-business-day clock after materiality determination |
| Deepfake/BEC or synthetic executive/vendor identity suspected | Legal, Finance, Fraud, Executive sponsor, Communications | Immediately; freeze payment/access action pending out-of-band verification |
| Insider threat involving executive or privileged admin | Legal counsel, HR, Board (if executive) | Immediately |
| IR team lacks expertise for the attack type | External IR retainer, Vendor support | Upon recognition |

Expand All @@ -367,7 +416,7 @@ Produce the incident response report with these exact sections:
```markdown
## Incident Response Report: [Incident ID]
**Date:** [YYYY-MM-DD]
**Skill:** ir-playbook v1.0.0
**Skill:** ir-playbook v1.0.2
**Frameworks:** NIST SP 800-61 Rev 2, SANS Incident Handler's Handbook
**Incident Commander:** [Name or "Unassigned -- assign immediately"]

Expand All @@ -386,6 +435,16 @@ and recommended immediate actions. Lead with the most critical fact.]
| Recoverability | [Regular / Supplemented / Extended / Not Recoverable] |
| Status | [Detected / Analyzing / Contained / Eradicated / Recovered / Closed] |

### Materiality and Privilege Handling
| Field | Value |
|---|---|
| Potentially Material | [Yes / No / Not Evaluable] |
| Materiality Owner | [Legal / Disclosure Committee / Executive / Missing] |
| Determination Timestamp | [timestamp / Pending / N/A] |
| SEC 4-Business-Day Clock | [Not started / Started / N/A / Not Evaluable] |
| Privilege Handling | [Counsel-led / Work product / Standard / Missing] |
| Evidence Gaps | [functional/financial/legal/customer/reputation gaps] |

### Timeline
| Timestamp (UTC) | Event | Source |
|---|---|---|
Expand All @@ -401,6 +460,21 @@ and recommended immediate actions. Lead with the most critical fact.]
|---|---|---|---|
| [Action taken] | [Complete / In Progress / Planned] | [timestamp] | [responder] |

### Cloud Identity and Exfiltration Controls
| Principal / Token / App | Risk | Containment Action | Evidence | Decision |
|---|---|---|---|---|
| [principal/token/app] | [refresh token/OAuth/API/service account] | [revoked/rotated/throttled] | [log/timestamp] | [Pass/Gap/Not Evaluable] |

### Wiper Backup Integrity Gate
| Backup / Restore Asset | Immutable | Malware Scan | Restore Test | Last Known Good | Decision |
|---|---|---|---|---|---|
| [backup set] | [yes/no/missing] | [pass/fail/missing] | [pass/fail/missing] | [timestamp] | [Pass/Gap/Not Evaluable] |

### Synthetic Social Engineering Verification
| Artifact | Requested Action | Out-of-Band Verification | Payment/Access Freeze | Decision |
|---|---|---|---|---|
| [voice/video/chat/email] | [payment/access/data] | [verified/failed/missing] | [yes/no] | [Pass/Gap/Not Evaluable] |

### Eradication and Recovery
- **Root Cause:** [Description of initial access vector and exploitation path]
- **Eradication Actions:** [List of removal actions taken]
Expand Down Expand Up @@ -468,6 +542,22 @@ Reconnecting systems to the network before thoroughly removing all persistence m

Breach notification regulations impose strict timelines that begin running at the moment of discovery, not at the conclusion of investigation. GDPR requires notification within 72 hours of becoming aware of a personal data breach. Missing these deadlines exposes the organization to regulatory penalties independent of the incident itself. Track notification deadlines from the moment a potential data breach is identified, and involve legal counsel early.

### Pitfall 6: Isolating a Host While Cloud Tokens Keep Working

Endpoint or subnet isolation does not revoke SaaS sessions, OAuth grants, API keys, cloud role sessions, or service-account credentials that the attacker already copied. Cloud-native exfiltration needs identity containment in parallel with network actions.

### Pitfall 7: Restoring Backups into a Wiper Loop

Wiper recovery can fail when backup images, startup scripts, GPOs, scheduled jobs, or backup agents were poisoned before the destructive payload triggered. Verify backup integrity and scan restored staging systems before production recovery.

### Pitfall 8: Treating Deepfake BEC as Ordinary Phishing

Synthetic voice/video incidents require payment or access freezes, out-of-band verification, preservation of media artifacts and call metadata, and finance/fraud escalation. Standard phishing triage can miss the business-process control failure.

### Pitfall 9: Over-Labeling Every Message as Privileged

Privilege handling protects legal advice and counsel-directed work product, but blanket labels on all incident messages can reduce credibility and confuse operations. For SEV-1 incidents, separate privileged legal analysis from operational coordination, record the counsel owner, and restrict distribution to need-to-know recipients.

---

## 8. Prompt Injection Safety Notice
Expand All @@ -491,9 +581,22 @@ This skill processes incident data that may include attacker-controlled content
5. **NIST Cybersecurity Framework (CSF) -- Respond Function** -- https://www.nist.gov/cyberframework
6. **GDPR Article 33** -- Notification of a personal data breach to the supervisory authority -- https://gdpr-info.eu/art-33-gdpr/
7. **HIPAA Breach Notification Rule** -- 45 CFR 164.400-414 -- https://www.hhs.gov/hipaa/for-professionals/breach-notification/
8. **SEC Cybersecurity Incident Disclosure (Item 1.05 Form 8-K)** -- https://www.sec.gov/rules/final/2023/33-11216.pdf
8. **SEC Cybersecurity Incident Disclosure (Item 1.05 Form 8-K)** -- https://www.sec.gov/newsroom/press-releases/2023-139
9. **FBI Internet Crime Complaint Center (IC3)** -- https://www.ic3.gov/
10. **FIRST CSIRT Framework** -- https://www.first.org/education/csirt
11. **CISA Destructive Malware Guidance** -- https://www.cisa.gov/topics/cyber-threats-and-advisories
12. **H-ISAC (Health Information Sharing and Analysis Center)** -- https://h-isac.org/
13. **KrebsOnSecurity: Iran-backed wiper attack on Stryker medtech (2026)** -- https://krebsonsystems.com/2026/03/iran-backed-hackers-claim-wiper-attack-on-medtech-firm-stryker/
14. **SEC Final Rule 33-11216 -- Cybersecurity Risk Management, Strategy, Governance, and Incident Disclosure** -- https://www.sec.gov/newsroom/press-releases/2023-139
15. **Microsoft Entra revoke user sign-in sessions** -- https://learn.microsoft.com/en-us/powershell/module/microsoft.graph.users.actions/revoke-mgusersigninsession
16. **CISA Secure by Demand Guide: deepfake and synthetic media risks** -- https://www.cisa.gov/resources-tools/resources/secure-demand-guide
17. **NIST AI RMF 1.0** -- https://www.nist.gov/itl/ai-risk-management-framework

---

## Version History

| Version | Date | Changes |
|---|---|---|
| 1.0.2 | 2026-06-06 | Add SEC materiality, cloud-token containment, wiper backup integrity, synthetic social engineering, and legal privilege evidence gates. |
| 1.0.1 | Initial | Add destructive malware response track. |
Original file line number Diff line number Diff line change
@@ -0,0 +1,133 @@
# SEC Materiality, Wiper, Cloud Token, and Deepfake Fixtures

These fixtures calibrate the supplemental incident-response gates in `SKILL.md`.

```yaml
case: sec_materiality_not_evaluable
incident:
company_type: public_company
category: data_exfiltration
affected_services:
- customer_portal
evidence:
functional_impact: partial
financial_impact: missing
legal_regulatory_impact: missing
customer_impact: missing
materiality_owner: missing
determination_timestamp: missing
expected_decision: Not Evaluable
expected_findings:
- check: IR-MAT-01
severity: High
reason: Potential materiality lacks financial, legal, customer, and reputational evidence.
- check: IR-MAT-02
severity: High
reason: SEC clock owner and materiality determination status are missing.
```

```yaml
case: executive_notice_without_privilege
incident:
severity: SEV-1
counsel_involved: true
notification:
audience: executive_leadership
privilege_header: missing
distribution: broad_all_managers
compromised_channel_assessed: false
expected_decision: Gap
expected_findings:
- check: IR-MAT-03
severity: Medium
reason: Counsel-led SEV-1 notification lacks privilege/work-product handling and need-to-know distribution.
```

```yaml
case: cloud_exfil_network_only
incident:
category: data_exfiltration
attacker_access:
oauth_refresh_token: present
service_account_key: present
containment:
endpoint_isolated: true
token_revocation: missing
oauth_grant_removal: missing
service_account_key_rotation: missing
api_log_validation: missing
expected_decision: Gap
expected_findings:
- check: IR-MAT-04
severity: Critical
reason: Valid cloud/SaaS tokens remain usable after endpoint isolation.
```

```yaml
case: wiper_restore_without_backup_integrity
incident:
category: destructive_wiper
recovery:
immutable_backup_status: missing
backup_malware_scan: missing
restore_test: missing
last_known_good: unknown
persistence_scan: missing
expected_decision: Not Evaluable
expected_findings:
- check: IR-MAT-05
severity: High
reason: Wiper recovery lacks immutable-backup, malware-scan, restore-test, and re-wipe-loop evidence.
```

```yaml
case: deepfake_bec_not_verified
incident:
category: social_engineering
artifact: executive_voice_call
requested_action: urgent_wire_transfer
evidence:
synthetic_media_review: missing
out_of_band_verification: missing
payment_freeze: false
call_metadata_preserved: false
expected_decision: Gap
expected_findings:
- check: IR-MAT-06
severity: High
reason: Deepfake/BEC indicators lack verification, payment freeze, and artifact preservation.
```

```yaml
case: complete_materiality_wiper_cloud_deepfake_package
incident:
severity: SEV-1
company_type: public_company
evidence:
materiality:
owner: disclosure_committee
functional_impact: documented
financial_impact: documented
legal_regulatory_impact: documented
customer_reputation_impact: documented
determination_timestamp: "2026-06-06T15:00:00Z"
privilege:
counsel_led: true
executive_notice_labeled: attorney_client_privilege_work_product
cloud_identity:
refresh_tokens_revoked: true
oauth_grants_removed: true
service_account_keys_rotated: true
api_logs_validated: true
wiper_recovery:
immutable_backup_verified: true
malware_scan_passed: true
restore_test_passed: true
last_known_good: "2026-06-05T00:00:00Z"
synthetic_social_engineering:
out_of_band_verification: passed
payment_freeze: true
artifact_preserved: true
expected_decision: Pass
expected_findings: []
```