external/anthropic-cybersecurity-skills/skills/building-soc-playbook-for-ransomware/SKILL.md
Builds a structured SOC incident response playbook for ransomware attacks covering detection, containment, eradication, and recovery phases with specific SIEM queries, isolation procedures, and decision trees. Use when SOC teams need formalized response procedures for ransomware incidents aligned to NIST SP 800-61 and MITRE ATT&CK ransomware techniques.
npx skillsauth add seikaikyo/dash-skills building-soc-playbook-for-ransomwareInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Use this skill when:
Do not use during an active ransomware incident as the sole guide — have pre-built playbooks tested and rehearsed before incidents occur.
Create SIEM detection rules for early ransomware indicators:
Mass File Encryption Detection (Splunk):
index=sysmon EventCode=11
| bin _time span=1m
| stats dc(TargetFilename) AS unique_files, values(TargetFilename) AS sample_files by Computer, Image, _time
| where unique_files > 100
| eval suspicious_extensions = if(match(mvjoin(sample_files, ","), "\.(encrypted|locked|crypt|enc|ransom)"), "YES", "NO")
| where suspicious_extensions="YES" OR unique_files > 500
| sort - unique_files
Shadow Copy Deletion (T1490):
index=wineventlog sourcetype="WinEventLog:Security" OR index=sysmon EventCode=1
(CommandLine="*vssadmin*delete*shadows*" OR CommandLine="*wmic*shadowcopy*delete*"
OR CommandLine="*bcdedit*/set*recoveryenabled*no*" OR CommandLine="*wbadmin*delete*catalog*")
| table _time, Computer, User, ParentImage, Image, CommandLine
Ransomware Note File Creation:
index=sysmon EventCode=11
TargetFilename IN ("*README*.txt", "*DECRYPT*.txt", "*RANSOM*.txt", "*RECOVER*.html", "*HOW_TO*.txt")
| stats count by Computer, Image, TargetFilename
| where count > 5
Elastic Security EQL variant:
sequence by host.name with maxspan=2m
[process where event.type == "start" and
process.args : ("*vssadmin*", "*delete*", "*shadows*")]
[file where event.type == "creation" and
file.name : ("*README*DECRYPT*", "*RANSOM*", "*HOW_TO_RECOVER*")]
RANSOMWARE ALERT TRIAGE
│
├── Is encryption actively occurring?
│ ├── YES → IMMEDIATE: Isolate host from network (Step 3)
│ │ Do NOT power off (preserve memory for forensics)
│ └── NO → Is this a pre-encryption indicator?
│ ├── Shadow copy deletion → HIGH PRIORITY: Isolate and investigate
│ ├── Known ransomware hash → HIGH PRIORITY: Block hash, scan enterprise
│ └── Suspicious process behavior → MEDIUM: Investigate, prepare isolation
│
├── How many hosts affected?
│ ├── Single host → Contained incident, follow host isolation procedure
│ ├── Multiple hosts (2-10) → Escalate to Tier 2, begin enterprise-wide scan
│ └── Enterprise-wide (>10) → Activate full IR team, engage external retainer
│
└── Is data exfiltration confirmed?
├── YES → Double extortion scenario, engage legal for breach notification
└── NO/UNKNOWN → Check for Cobalt Strike/C2 beacons, review outbound transfers
Network Isolation via EDR (CrowdStrike Falcon):
# Isolate host using CrowdStrike Falcon API
curl -X POST "https://api.crowdstrike.com/devices/entities/devices-actions/v2?action_name=contain" \
-H "Authorization: Bearer $TOKEN" \
-H "Content-Type: application/json" \
-d '{"ids": ["device_id_here"]}'
Network Isolation via Microsoft Defender for Endpoint:
# Isolate machine via MDE API
$headers = @{Authorization = "Bearer $token"}
$body = @{Comment = "Ransomware containment - IR-2024-0500"; IsolationType = "Full"} | ConvertTo-Json
Invoke-RestMethod -Uri "https://api.securitycenter.microsoft.com/api/machines/$machineId/isolate" `
-Method Post -Headers $headers -Body $body -ContentType "application/json"
Firewall Emergency Rules:
# Palo Alto — Block SMB lateral spread
set rulebase security rules RansomwareContainment from Trust to Trust
set rulebase security rules RansomwareContainment application ms-ds-smb
set rulebase security rules RansomwareContainment action deny
set rulebase security rules RansomwareContainment disabled no
commit
Active Directory Emergency Actions:
# Disable compromised account
Disable-ADAccount -Identity "compromised_user"
# Reset Kerberos TGT (if domain admin compromised)
# WARNING: This resets krbtgt and requires two resets 12+ hours apart
Reset-KrbtgtKeys -Server "DC-PRIMARY" -Force
# Block lateral movement by disabling remote services
Set-Service -Name "RemoteRegistry" -StartupType Disabled -Status Stopped
Collect forensic artifacts before remediation:
# Capture running processes and network connections
Get-Process | Export-Csv "C:\IR\processes_$(hostname).csv"
Get-NetTCPConnection | Export-Csv "C:\IR\netstat_$(hostname).csv"
# Capture memory dump (if host still running)
winpmem_mini_x64.exe C:\IR\memory_$(hostname).raw
# Collect ransomware artifacts
Copy-Item "C:\Users\*\Desktop\*README*" "C:\IR\ransom_notes\" -Recurse
Copy-Item "C:\Users\*\Desktop\*.encrypted" "C:\IR\encrypted_samples\" -Force
# Capture event logs
wevtutil epl Security "C:\IR\Security_$(hostname).evtx"
wevtutil epl System "C:\IR\System_$(hostname).evtx"
wevtutil epl "Microsoft-Windows-Sysmon/Operational" "C:\IR\Sysmon_$(hostname).evtx"
Identify ransomware variant:
Enterprise-wide IOC scan in Splunk:
index=sysmon (EventCode=1 OR EventCode=11 OR EventCode=3)
(TargetFilename="*ransomware_binary_name*" OR sha256="KNOWN_HASH"
OR DestinationIp="C2_IP_ADDRESS" OR CommandLine="*malicious_command*")
| stats count by Computer, EventCode, Image, CommandLine
| sort - count
Recovery from backups:
Structure the playbook conclusion with lessons learned:
POST-INCIDENT REVIEW TEMPLATE
1. Timeline of events (detection to full recovery)
2. Initial access vector identification
3. Dwell time analysis (time from initial compromise to encryption)
4. Detection gaps identified
5. Response effectiveness metrics (MTTD, MTTC, MTTR)
6. Playbook improvements recommended
7. New detection rules deployed
8. Backup and recovery procedure updates
| Term | Definition | |------|-----------| | Double Extortion | Ransomware tactic combining data encryption with data theft, threatening public release if ransom unpaid | | Dwell Time | Duration between initial compromise and detection — ransomware operators average 5-9 days before encryption | | MTTC | Mean Time to Contain — time from detection to successful isolation of affected systems | | Kill Chain | Ransomware progression: Initial Access -> Execution -> Persistence -> Privilege Escalation -> Lateral Movement -> Collection -> Exfiltration -> Impact | | Immutable Backup | Backup storage that cannot be modified or deleted for a defined retention period (WORM storage) | | RTO/RPO | Recovery Time Objective / Recovery Point Objective — maximum acceptable downtime and data loss thresholds |
RANSOMWARE PLAYBOOK EXECUTION — IR-2024-0500
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Phase 1 - Detection:
Alert: Mass file encryption detected on FILESERVER-03
Variant: LockBit 3.0 (confirmed via ID Ransomware)
MTTD: 12 minutes from first encryption to SOC alert
Phase 2 - Containment:
[DONE] FILESERVER-03 isolated via CrowdStrike at 14:35 UTC
[DONE] SMB blocked enterprise-wide via firewall emergency rule
[DONE] Compromised service account disabled in AD
MTTC: 23 minutes
Phase 3 - Eradication:
[DONE] 3 additional hosts with C2 beacon identified and isolated
[DONE] Cobalt Strike C2 domain (c2[.]evil[.]com) sinkholed
[DONE] Enterprise-wide IOC scan completed — no additional infections
Phase 4 - Recovery:
[DONE] FILESERVER-03 rebuilt from gold image
[DONE] Data restored from immutable Veeam backup (RPO: 4 hours)
[DONE] Systems monitored 72 hours — no reinfection
MTTR: 18 hours
Total Affected: 1 server, 3 workstations
Data Loss: 4 hours of file modifications (backup RPO)
Exfiltration: No evidence of data exfiltration confirmed
development
拋棄式 HTML mockup 比稿:產出 2 到 3 個設計立場不同的變體(密度 / 版式 / 強調軸,不是換色),各附取捨說明,最後給有立場的對比結論。適用:「畫個草圖」「比較 A 版 B 版」「先看方向再做」「給我看幾種做法」。要 production 元件或設計已定案時不適用。
tools
需求不明時的意圖萃取訪談:一次一題、每題附上自己的猜測、聽出「真正想要 vs 覺得應該要」,直到能預測使用者反應(約 95% 信心)才動工。適用:需求缺少對象 / 動機 / 成功標準 / 約束,或使用者點名「訪談我」「先確認一下」「我們確定嗎」。明確自足的指示、純資訊查詢、機械性操作不適用。
development
對非平凡決策啟動新鮮 context 對抗審查(找碴不背書),在修正還便宜的時候抓出錯誤方向。適用:高風險改動(production、資安敏感邏輯、不可逆操作)、不熟的程式碼、要宣稱「這樣是安全的 / 可行的」之前。機械性操作與一行修改不適用。
testing
Reference for writing and editing agent skills well — the vocabulary and principles that make a skill predictable. Consult when authoring, reviewing, or pruning a SKILL.md.