npx skills add mukul975/Anthropic-Cybersecurity-SkillsMITRE ATT&CK
When to Use
Use this skill when:
- Building a SOAR playbook that automatically enriches SIEM alerts with threat intelligence context before routing to analysts
- Creating a Python pipeline for bulk IOC enrichment from phishing email submissions
- Reducing analyst mean time to triage (MTTT) by pre-populating alert context with VT, Shodan, and MISP data
Do not use this skill for fully automated blocking decisions without human review — enrichment automation should inform decisions, not execute blocks autonomously for high-impact actions.
Prerequisites
- SOAR platform (Cortex XSOAR, Splunk SOAR, Tines, or n8n) or Python 3.9+ environment
- API keys: VirusTotal, AbuseIPDB, Shodan, and at minimum one TIP (MISP or OpenCTI)
- SIEM integration endpoint for alert consumption
- Rate limit budgets documented per API (VT: 4/min free, 500/min enterprise)
Workflow
Step 1: Design Enrichment Pipeline Architecture
Define the enrichment flow for each IOC type:
SIEM Alert → Extract IOCs → Classify Type → Route to enrichment functions
IP Address → AbuseIPDB + Shodan + VirusTotal IP + MISP
Domain → VirusTotal Domain + PassiveTotal + Shodan + MISP
URL → URLScan.io + VirusTotal URL + Google Safe Browse
File Hash → VirusTotal Files + MalwareBazaar + MISP
→ Aggregate results → Calculate confidence score → Update alert → Notify analystStep 2: Implement Python Enrichment Functions
import requests
import time
from dataclasses import dataclass, field
from typing import Optional
RATE_LIMIT_DELAY = 0.25 # 4 requests/second for VT free tier
@dataclass
class EnrichmentResult:
ioc_value: str
ioc_type: str
vt_malicious: int = 0
vt_total: int = 0
abuse_confidence: int = 0
shodan_ports: list = field(default_factory=list)
misp_events: list = field(default_factory=list)
confidence_score: int = 0
def enrich_ip(ip: str, vt_key: str, abuse_key: str, shodan_key: str) -> EnrichmentResult:
result = EnrichmentResult(ip, "ip")
# VirusTotal IP lookup
vt_resp = requests.get(
f"https://www.virustotal.com/api/v3/ip_addresses/{ip}",
headers={"x-apikey": vt_key}
)
if vt_resp.status_code == 200:
stats = vt_resp.json()["data"]["attributes"]["last_analysis_stats"]
result.vt_malicious = stats.get("malicious", 0)
result.vt_total = sum(stats.values())
time.sleep(RATE_LIMIT_DELAY)
# AbuseIPDB
abuse_resp = requests.get(
"https://api.abuseipdb.com/api/v2/check",
headers={"Key": abuse_key, "Accept": "application/json"},
params={"ipAddress": ip, "maxAgeInDays": 90}
)
if abuse_resp.status_code == 200:
result.abuse_confidence = abuse_resp.json()["data"]["abuseConfidenceScore"]
# Calculate composite confidence score
result.confidence_score = min(
(result.vt_malicious / max(result.vt_total, 1)) * 60 +
(result.abuse_confidence / 100) * 40, 100
)
return result
def enrich_hash(sha256: str, vt_key: str) -> EnrichmentResult:
result = EnrichmentResult(sha256, "sha256")
vt_resp = requests.get(
f"https://www.virustotal.com/api/v3/files/{sha256}",
headers={"x-apikey": vt_key}
)
if vt_resp.status_code == 200:
stats = vt_resp.json()["data"]["attributes"]["last_analysis_stats"]
result.vt_malicious = stats.get("malicious", 0)
result.vt_total = sum(stats.values())
result.confidence_score = int((result.vt_malicious / max(result.vt_total, 1)) * 100)
return resultStep 3: Build SOAR Playbook (Cortex XSOAR)
In Cortex XSOAR, create an enrichment playbook:
- Trigger: Alert created in SIEM (via webhook or polling)
- Extract IOCs: Use "Extract Indicators" task with regex patterns for IP, domain, URL, hash
- Parallel enrichment: Fan-out to multiple enrichment tasks simultaneously
- VT Enrichment: Call
!vt-file-scanor!vt-ip-scancommands - AbuseIPDB check: Call
!abuseipdb-check-ipcommand - MISP Lookup: Call
!misp-searchfor cross-referencing - Score aggregation: Python transform task computing composite score
- Conditional routing: If score ≥70 → High Priority queue; if 40–69 → Medium; <40 → Auto-close with note
- Alert enrichment: Write enrichment results to alert context for analyst view
Step 4: Handle Rate Limiting and Failures
import time
from functools import wraps
def rate_limited(max_per_second):
min_interval = 1.0 / max_per_second
def decorator(func):
last_called = [0.0]
@wraps(func)
def wrapper(*args, **kwargs):
elapsed = time.time() - last_called[0]
wait = min_interval - elapsed
if wait > 0:
time.sleep(wait)
result = func(*args, **kwargs)
last_called[0] = time.time()
return result
return wrapper
return decorator
def retry_on_429(max_retries=3):
def decorator(func):
@wraps(func)
def wrapper(*args, **kwargs):
for attempt in range(max_retries):
response = func(*args, **kwargs)
if response.status_code == 429:
retry_after = int(response.headers.get("Retry-After", 60))
time.sleep(retry_after)
else:
return response
return wrapper
return decoratorStep 5: Metrics and Tuning
Track pipeline performance weekly:
- Enrichment latency: Target <30 seconds from alert trigger to enriched output
- API success rate: Target >99% (identify rate limit or outage events)
- True positive rate: Track analyst overrides of automated confidence scores
- Cost: Track API call volume against budget (VT Enterprise: $X per 1M lookups)
Key Concepts
| Term | Definition |
|---|---|
| SOAR | Security Orchestration, Automation, and Response — platform for automating security workflows and integrating disparate tools |
| Enrichment Playbook | Automated workflow sequence that adds contextual intelligence to raw security events |
| Rate Limiting | API provider restrictions on request frequency (e.g., VT free: 4 requests/minute); pipelines must respect these limits |
| Composite Confidence Score | Single score aggregating signals from multiple enrichment sources using weighted formula |
| Fan-out Pattern | Parallel execution of multiple enrichment queries simultaneously to minimize total enrichment latency |
Tools & Systems
- Cortex XSOAR (Palo Alto): Enterprise SOAR with 700+ marketplace integrations including VT, MISP, Shodan, and AbuseIPDB
- Splunk SOAR (Phantom): SOAR platform with Python-based playbooks; native Splunk SIEM integration
- Tines: No-code SOAR platform with webhook-driven automation; cost-effective for smaller teams
- TheHive + Cortex: Open-source IR/enrichment platform with observable enrichment via Cortex analyzers
Common Pitfalls
- Blocking on enrichment latency: If enrichment takes >5 minutes, analysts start working unenriched alerts, defeating the purpose. Set timeout limits and provide partial results.
- No caching: Querying the same IOC 50 times generates unnecessary API costs. Cache enrichment results for 24 hours by default.
- Ignoring API failures silently: Failed enrichment calls should be logged and trigger fallback logic, not silently produce empty results that appear as clean IOCs.
- Automating blocks on enrichment score alone: Composite scores contain false positives; require human confirmation for blocking decisions against shared infrastructure.
References and resources
Everything below is rendered for inspection. Script files are read-only and never run.
References 1
api-reference.md1.8 KB
API Reference: Automating IOC Enrichment
VirusTotal API v3
IP Lookup
import requests
resp = requests.get(
"https://www.virustotal.com/api/v3/ip_addresses/1.2.3.4",
headers={"x-apikey": VT_KEY},
)
stats = resp.json()["data"]["attributes"]["last_analysis_stats"]
print(stats["malicious"], "/", sum(stats.values()))File Hash Lookup
resp = requests.get(
f"https://www.virustotal.com/api/v3/files/{sha256}",
headers={"x-apikey": VT_KEY},
)Domain Lookup
resp = requests.get(
f"https://www.virustotal.com/api/v3/domains/{domain}",
headers={"x-apikey": VT_KEY},
)AbuseIPDB API v2
resp = requests.get(
"https://api.abuseipdb.com/api/v2/check",
headers={"Key": ABUSE_KEY, "Accept": "application/json"},
params={"ipAddress": "1.2.3.4", "maxAgeInDays": 90},
)
data = resp.json()["data"]
print("Confidence:", data["abuseConfidenceScore"])
print("Reports:", data["totalReports"])Shodan API
import shodan
api = shodan.Shodan(SHODAN_KEY)
info = api.host("1.2.3.4")
print("Ports:", info.get("ports"))
print("Vulns:", info.get("vulns"))STIX 2.1 Export
from stix2 import Indicator, Bundle
indicator = Indicator(
pattern="[ipv4-addr:value = '1.2.3.4']",
pattern_type="stix",
valid_from="2025-01-01T00:00:00Z",
confidence=85,
)
bundle = Bundle(objects=[indicator])Rate Limits
| API | Free Tier | Enterprise |
|---|---|---|
| VirusTotal | 4 req/min | 500 req/min |
| AbuseIPDB | 1000 req/day | 5000 req/day |
| Shodan | 1 req/sec | 10 req/sec |
References
- VirusTotal API: https://docs.virustotal.com/reference/overview
- AbuseIPDB API: https://docs.abuseipdb.com/
- stix2 library: https://pypi.org/project/stix2/
- Shodan: https://shodan.readthedocs.io/
Scripts 1
agent.py7.8 KB
#!/usr/bin/env python3
"""Agent for automating IOC enrichment with VirusTotal, AbuseIPDB, and STIX."""
import os
import re
import json
import time
import argparse
from datetime import datetime
from dataclasses import dataclass, field
import requests
from stix2 import Indicator, Bundle
RATE_LIMIT_DELAY = 0.25
@dataclass
class EnrichmentResult:
ioc_value: str
ioc_type: str
vt_malicious: int = 0
vt_total: int = 0
vt_threat_label: str = ""
abuse_confidence: int = 0
abuse_reports: int = 0
shodan_ports: list = field(default_factory=list)
confidence_score: int = 0
def classify_ioc(value):
"""Auto-detect IOC type from value."""
if re.match(r"^\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}$", value):
return "ip"
if re.match(r"^[a-fA-F0-9]{64}$", value):
return "sha256"
if re.match(r"^[a-fA-F0-9]{32}$", value):
return "md5"
if re.match(r"^https?://", value):
return "url"
return "domain"
def enrich_ip_virustotal(ip, api_key):
"""Enrich an IP address via VirusTotal API v3."""
resp = requests.get(
f"https://www.virustotal.com/api/v3/ip_addresses/{ip}",
headers={"x-apikey": api_key},
timeout=30,
)
if resp.status_code == 200:
attrs = resp.json()["data"]["attributes"]
stats = attrs.get("last_analysis_stats", {})
return {
"malicious": stats.get("malicious", 0),
"total": sum(stats.values()),
"country": attrs.get("country", ""),
"asn": attrs.get("asn", 0),
"as_owner": attrs.get("as_owner", ""),
}
return {}
def enrich_hash_virustotal(file_hash, api_key):
"""Enrich a file hash via VirusTotal API v3."""
resp = requests.get(
f"https://www.virustotal.com/api/v3/files/{file_hash}",
headers={"x-apikey": api_key},
timeout=30,
)
if resp.status_code == 200:
attrs = resp.json()["data"]["attributes"]
stats = attrs.get("last_analysis_stats", {})
ptc = attrs.get("popular_threat_classification", {})
return {
"malicious": stats.get("malicious", 0),
"total": sum(stats.values()),
"threat_label": ptc.get("suggested_threat_label", ""),
"type_description": attrs.get("type_description", ""),
}
return {}
def enrich_domain_virustotal(domain, api_key):
"""Enrich a domain via VirusTotal API v3."""
resp = requests.get(
f"https://www.virustotal.com/api/v3/domains/{domain}",
headers={"x-apikey": api_key},
timeout=30,
)
if resp.status_code == 200:
attrs = resp.json()["data"]["attributes"]
stats = attrs.get("last_analysis_stats", {})
return {
"malicious": stats.get("malicious", 0),
"total": sum(stats.values()),
"registrar": attrs.get("registrar", ""),
}
return {}
def enrich_ip_abuseipdb(ip, api_key):
"""Check an IP against AbuseIPDB."""
resp = requests.get(
"https://api.abuseipdb.com/api/v2/check",
headers={"Key": api_key, "Accept": "application/json"},
params={"ipAddress": ip, "maxAgeInDays": 90},
timeout=30,
)
if resp.status_code == 200:
data = resp.json()["data"]
return {
"abuse_confidence": data.get("abuseConfidenceScore", 0),
"total_reports": data.get("totalReports", 0),
"country": data.get("countryCode", ""),
"isp": data.get("isp", ""),
}
return {}
def compute_confidence(vt_result, abuse_result=None):
"""Calculate composite confidence score from enrichment sources."""
vt_score = 0
if vt_result.get("total", 0) > 0:
vt_score = (vt_result["malicious"] / vt_result["total"]) * 60
abuse_score = 0
if abuse_result:
abuse_score = (abuse_result.get("abuse_confidence", 0) / 100) * 40
return min(int(vt_score + abuse_score), 100)
def enrich_ioc(value, ioc_type, vt_key, abuse_key=None):
"""Enrich a single IOC through all available sources."""
result = EnrichmentResult(ioc_value=value, ioc_type=ioc_type)
vt_data = {}
if ioc_type == "ip":
vt_data = enrich_ip_virustotal(value, vt_key)
time.sleep(RATE_LIMIT_DELAY)
if abuse_key:
abuse_data = enrich_ip_abuseipdb(value, abuse_key)
result.abuse_confidence = abuse_data.get("abuse_confidence", 0)
result.abuse_reports = abuse_data.get("total_reports", 0)
elif ioc_type in ("sha256", "md5"):
vt_data = enrich_hash_virustotal(value, vt_key)
time.sleep(RATE_LIMIT_DELAY)
elif ioc_type == "domain":
vt_data = enrich_domain_virustotal(value, vt_key)
time.sleep(RATE_LIMIT_DELAY)
result.vt_malicious = vt_data.get("malicious", 0)
result.vt_total = vt_data.get("total", 0)
result.vt_threat_label = vt_data.get("threat_label", "")
abuse_dict = {"abuse_confidence": result.abuse_confidence} if ioc_type == "ip" else None
result.confidence_score = compute_confidence(vt_data, abuse_dict)
return result
def export_stix_indicators(results, output_path):
"""Export enriched IOCs as STIX 2.1 indicators."""
pattern_map = {
"ip": lambda v: f"[ipv4-addr:value = '{v}']",
"domain": lambda v: f"[domain-name:value = '{v}']",
"sha256": lambda v: f"[file:hashes.'SHA-256' = '{v}']",
"md5": lambda v: f"[file:hashes.MD5 = '{v}']",
"url": lambda v: f"[url:value = '{v}']",
}
indicators = []
for r in results:
pattern_fn = pattern_map.get(r.ioc_type)
if pattern_fn:
ind = Indicator(
name=f"{r.ioc_type}: {r.ioc_value}",
pattern=pattern_fn(r.ioc_value),
pattern_type="stix",
valid_from=datetime.utcnow().strftime("%Y-%m-%dT%H:%M:%SZ"),
confidence=r.confidence_score,
)
indicators.append(ind)
bundle = Bundle(objects=indicators, allow_custom=True)
with open(output_path, "w") as f:
f.write(bundle.serialize(pretty=True))
return len(indicators)
def main():
parser = argparse.ArgumentParser(description="IOC Enrichment Automation Agent")
parser.add_argument("--vt-key", default=os.getenv("VT_API_KEY"), help="VirusTotal API key")
parser.add_argument("--abuse-key", default=os.getenv("ABUSEIPDB_KEY"), help="AbuseIPDB API key")
parser.add_argument("--ioc-file", help="File with IOCs (one per line)")
parser.add_argument("--ioc", help="Single IOC to enrich")
parser.add_argument("--output", default="enrichment_results.json")
parser.add_argument("--stix-output", help="Export as STIX bundle")
args = parser.parse_args()
iocs = []
if args.ioc:
iocs.append(args.ioc)
if args.ioc_file:
with open(args.ioc_file) as f:
iocs.extend(line.strip() for line in f if line.strip() and not line.startswith("#"))
results = []
for ioc_val in iocs:
ioc_type = classify_ioc(ioc_val)
print(f" Enriching {ioc_type}: {ioc_val}...")
result = enrich_ioc(ioc_val, ioc_type, args.vt_key, args.abuse_key)
results.append(result)
verdict = "MALICIOUS" if result.confidence_score >= 70 else "SUSPICIOUS" if result.confidence_score >= 40 else "CLEAN"
print(f" VT: {result.vt_malicious}/{result.vt_total} | Confidence: {result.confidence_score} | {verdict}")
report = {
"enriched_at": datetime.utcnow().isoformat(),
"total_iocs": len(results),
"results": [r.__dict__ for r in results],
}
with open(args.output, "w") as f:
json.dump(report, f, indent=2, default=str)
print(f"[+] Results saved to {args.output}")
if args.stix_output:
count = export_stix_indicators(results, args.stix_output)
print(f"[+] Exported {count} STIX indicators to {args.stix_output}")
if __name__ == "__main__":
main()