external/anthropic-cybersecurity-skills/skills/implementing-purdue-model-network-segmentation/SKILL.md
Implement network segmentation based on the Purdue Enterprise Reference Architecture (PERA) model, separating ICS networks into hierarchical security zones from Level 0 physical process through Level 5 enterprise and enforcing strict traffic control through IEC 62443-aligned DMZs between OT and IT domains. Use when designing ICS/SCADA network zones or segmenting OT from IT networks.
npx skillsauth add seikaikyo/dash-skills implementing-purdue-model-network-segmentationInstall 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.
Do not use for micro-segmentation within a single Purdue level (see implementing-zone-conduit-model-for-ics), for cloud-native environments without traditional ICS networks, or for network segmentation in purely IT environments.
Classify all network assets and data flows according to the Purdue Model hierarchy.
#!/usr/bin/env python3
"""Purdue Model Network Segmentation Planner.
Maps existing OT/IT network assets to Purdue Model levels and generates
segmentation recommendations including firewall rules and VLAN assignments.
"""
import json
import csv
import sys
from collections import defaultdict
from datetime import datetime
from typing import Dict, List
PURDUE_LEVELS = {
0: {
"name": "Physical Process",
"description": "Sensors, actuators, field instruments",
"typical_devices": ["Sensors", "Actuators", "Drives", "Motor starters"],
"vlan_range": "100-109",
"allowed_protocols": ["HART", "Profibus", "Foundation Fieldbus", "IO-Link"],
},
1: {
"name": "Basic Control",
"description": "PLCs, RTUs, safety controllers",
"typical_devices": ["PLC", "RTU", "Safety Controller", "DCS Controller"],
"vlan_range": "110-119",
"allowed_protocols": ["EtherNet/IP", "Profinet", "Modbus/TCP", "S7comm", "OPC UA"],
},
2: {
"name": "Supervisory Control",
"description": "HMI, SCADA servers, engineering workstations",
"typical_devices": ["HMI", "SCADA Server", "Engineering Workstation", "Batch Server"],
"vlan_range": "120-129",
"allowed_protocols": ["OPC UA", "OPC DA", "Modbus/TCP", "DNP3", "HTTPS"],
},
3: {
"name": "Site Operations",
"description": "Historian, MES, asset management",
"typical_devices": ["Historian", "MES Server", "Asset Management", "Alarm Server"],
"vlan_range": "130-139",
"allowed_protocols": ["OPC UA", "SQL", "HTTPS", "MQTT"],
},
3.5: {
"name": "IT/OT DMZ",
"description": "Demilitarized zone between IT and OT",
"typical_devices": ["Jump Server", "Historian Mirror", "Patch Server", "AV Update Server", "Remote Access Gateway"],
"vlan_range": "150-159",
"allowed_protocols": ["HTTPS", "RDP (to jump server only)", "SSH", "SQL (read replica)"],
},
4: {
"name": "Enterprise IT",
"description": "Enterprise applications, email, ERP",
"typical_devices": ["ERP Server", "Email Server", "Business Applications", "Active Directory"],
"vlan_range": "200-249",
"allowed_protocols": ["HTTPS", "LDAPS", "SMTP", "SQL"],
},
5: {
"name": "Enterprise Network / Internet",
"description": "External connections, cloud services, partner networks",
"typical_devices": ["Internet Gateway", "VPN Concentrator", "Cloud Services"],
"vlan_range": "250-254",
"allowed_protocols": ["HTTPS", "IPsec VPN"],
},
}
class PurdueSegmentationPlanner:
"""Plans Purdue Model network segmentation."""
def __init__(self):
self.assets = []
self.data_flows = []
self.firewall_rules = []
def load_asset_inventory(self, filepath: str):
"""Load asset inventory from CSV."""
with open(filepath, "r") as f:
self.assets = list(csv.DictReader(f))
print(f"[*] Loaded {len(self.assets)} assets")
def classify_assets(self):
"""Classify assets into Purdue levels based on type and function."""
classification = defaultdict(list)
for asset in self.assets:
level = asset.get("purdue_level", "")
try:
level = float(level)
except (ValueError, TypeError):
level = self._infer_purdue_level(asset)
classification[level].append(asset)
asset["assigned_level"] = level
return classification
def _infer_purdue_level(self, asset: dict) -> float:
"""Infer Purdue level from device type if not explicitly assigned."""
device_type = asset.get("type", "").lower()
mapping = {
"sensor": 0, "actuator": 0, "drive": 0,
"plc": 1, "rtu": 1, "safety": 1, "dcs": 1,
"hmi": 2, "scada": 2, "engineering": 2,
"historian": 3, "mes": 3, "alarm": 3,
"jump": 3.5, "patch": 3.5, "remote_access": 3.5,
"erp": 4, "email": 4, "directory": 4,
}
for keyword, level in mapping.items():
if keyword in device_type:
return level
return -1
def generate_vlan_plan(self, classification: dict) -> list:
"""Generate VLAN assignment plan based on Purdue levels."""
vlan_plan = []
for level, info in PURDUE_LEVELS.items():
assets_at_level = classification.get(level, [])
vlan_plan.append({
"purdue_level": level,
"level_name": info["name"],
"vlan_range": info["vlan_range"],
"asset_count": len(assets_at_level),
"allowed_protocols": info["allowed_protocols"],
})
return vlan_plan
def generate_firewall_rules(self) -> list:
"""Generate inter-level firewall rules enforcing Purdue Model boundaries."""
rules = [
{
"rule_id": 1,
"name": "Block direct IT-to-Level1",
"action": "DENY",
"source_zone": "Level_4_Enterprise",
"dest_zone": "Level_1_Control",
"service": "ANY",
"log": True,
"description": "No direct access from enterprise IT to basic control PLCs",
},
{
"rule_id": 2,
"name": "Block direct IT-to-Level2",
"action": "DENY",
"source_zone": "Level_4_Enterprise",
"dest_zone": "Level_2_Supervisory",
"service": "ANY",
"log": True,
"description": "No direct access from enterprise IT to HMI/SCADA",
},
{
"rule_id": 3,
"name": "Allow DMZ-to-Historian-Replica",
"action": "ALLOW",
"source_zone": "Level_3_Operations",
"dest_zone": "Level_35_DMZ",
"service": "SQL/1433 (read replica push)",
"log": True,
"description": "Historian pushes data to DMZ replica for IT consumption",
},
{
"rule_id": 4,
"name": "Allow IT-to-DMZ-JumpServer",
"action": "ALLOW",
"source_zone": "Level_4_Enterprise",
"dest_zone": "Level_35_DMZ",
"service": "RDP/3389, SSH/22",
"log": True,
"description": "IT users access OT via jump server in DMZ only",
},
{
"rule_id": 5,
"name": "Allow DMZ-JumpServer-to-Level2",
"action": "ALLOW",
"source_zone": "Level_35_DMZ",
"dest_zone": "Level_2_Supervisory",
"service": "RDP/3389 (from jump server IP only)",
"log": True,
"description": "Jump server provides controlled access to HMI/SCADA",
},
{
"rule_id": 6,
"name": "Allow Level2-to-Level1",
"action": "ALLOW",
"source_zone": "Level_2_Supervisory",
"dest_zone": "Level_1_Control",
"service": "Modbus/502, EtherNet-IP/44818, S7comm/102",
"log": True,
"description": "HMI/SCADA communicates with PLCs using industrial protocols",
},
{
"rule_id": 7,
"name": "Block Level1-outbound-internet",
"action": "DENY",
"source_zone": "Level_1_Control",
"dest_zone": "Level_5_Internet",
"service": "ANY",
"log": True,
"description": "PLCs must never reach the internet directly",
},
{
"rule_id": 8,
"name": "Allow patch distribution DMZ-to-Level2",
"action": "ALLOW",
"source_zone": "Level_35_DMZ",
"dest_zone": "Level_2_Supervisory",
"service": "WSUS/8530",
"log": True,
"description": "Patch server in DMZ distributes updates to supervisory systems",
},
{
"rule_id": 9,
"name": "Default deny all inter-zone",
"action": "DENY",
"source_zone": "ANY",
"dest_zone": "ANY",
"service": "ANY",
"log": True,
"description": "Default deny all traffic not explicitly permitted",
},
]
self.firewall_rules = rules
return rules
def print_segmentation_plan(self, classification: dict):
"""Print the complete segmentation plan."""
print(f"\n{'='*70}")
print("PURDUE MODEL NETWORK SEGMENTATION PLAN")
print(f"{'='*70}")
print(f"Generated: {datetime.now().isoformat()}")
vlan_plan = self.generate_vlan_plan(classification)
print(f"\n--- VLAN ASSIGNMENT ---")
for v in vlan_plan:
print(f"\n {v['level_name']} (Purdue {v['purdue_level']})")
print(f" VLAN Range: {v['vlan_range']}")
print(f" Assets: {v['asset_count']}")
print(f" Allowed Protocols: {', '.join(v['allowed_protocols'])}")
print(f"\n--- INTER-ZONE FIREWALL RULES ---")
rules = self.generate_firewall_rules()
for rule in rules:
action_symbol = "+" if rule["action"] == "ALLOW" else "X"
print(f"\n [{action_symbol}] Rule {rule['rule_id']}: {rule['name']}")
print(f" {rule['source_zone']} -> {rule['dest_zone']}")
print(f" Service: {rule['service']}")
print(f" Reason: {rule['description']}")
if __name__ == "__main__":
planner = PurdueSegmentationPlanner()
if len(sys.argv) >= 2:
planner.load_asset_inventory(sys.argv[1])
classification = planner.classify_assets()
planner.print_segmentation_plan(classification)
The DMZ is the critical boundary between IT and OT. All data exchange must traverse it -- no direct connections are permitted.
# Level 3.5 DMZ Architecture Configuration
# All IT-OT data exchange flows through the DMZ
dmz_architecture:
zone_name: "IT_OT_DMZ"
purdue_level: 3.5
vlan: 150
components:
historian_replica:
purpose: "Read-only copy of OT historian data for IT/business access"
direction: "OT pushes data TO DMZ (unidirectional)"
ip: "10.10.150.10"
services:
- port: 1433
protocol: "SQL"
direction: "inbound from Level 3 historian only"
- port: 443
protocol: "HTTPS"
direction: "outbound to Level 4 for IT consumers"
jump_server:
purpose: "Controlled remote access point for OT maintenance"
ip: "10.10.150.20"
services:
- port: 3389
protocol: "RDP"
direction: "inbound from Level 4 with MFA"
- port: 3389
protocol: "RDP"
direction: "outbound to Level 2 HMIs only"
security_controls:
- "Multi-factor authentication required"
- "Session recording enabled"
- "Maximum session duration: 4 hours"
- "Approval-based access workflow"
patch_server:
purpose: "Staging area for tested patches before OT deployment"
ip: "10.10.150.30"
services:
- port: 8530
protocol: "WSUS"
direction: "pulls from Level 4 WSUS, pushes to Level 2-3"
antivirus_relay:
purpose: "AV signature distribution to OT endpoints"
ip: "10.10.150.40"
services:
- port: 443
protocol: "HTTPS"
direction: "pulls definitions from Level 4, distributes to Level 2-3"
firewall_rules:
north_firewall: # Between DMZ and Level 4 Enterprise
- allow: "Level 4 -> DMZ jump server:3389 (with MFA)"
- allow: "Level 4 -> DMZ historian replica:443 (read-only)"
- allow: "DMZ patch server -> Level 4 WSUS:8530 (pull only)"
- deny: "ALL other traffic"
south_firewall: # Between DMZ and Level 3 Operations
- allow: "Level 3 historian -> DMZ replica:1433 (push direction)"
- allow: "DMZ jump server -> Level 2 HMI:3389 (session-limited)"
- allow: "DMZ patch server -> Level 2/3:8530 (scheduled)"
- deny: "ALL other traffic"
critical_rule: "NO traffic passes through DMZ end-to-end. DMZ breaks all connections."
| Term | Definition | |------|------------| | Purdue Model (PERA) | Hierarchical reference architecture organizing industrial networks into levels 0-5 based on function and trust | | Level 3.5 DMZ | Demilitarized zone between IT (Level 4) and OT (Level 3), where all cross-boundary data exchange occurs | | Defense in Depth | Layered security approach requiring attackers to breach multiple boundaries to reach critical control systems | | Data Diode | Hardware-enforced unidirectional communication device ensuring data flows only from OT to IT, never reverse | | Zone | Logical grouping of assets sharing common security requirements as defined by IEC 62443 | | Conduit | Controlled communication path between zones with defined security policies |
Context: An audit reveals that enterprise IT systems can directly communicate with PLCs on the control network. There is no DMZ and no firewall between IT and OT.
Approach:
Pitfalls: Do not cut over all traffic simultaneously -- migrate flow by flow with rollback plans. Legacy OT systems may use protocols that cannot traverse firewalls doing DPI; test thoroughly in a lab first. Never deploy the DMZ during active production without an agreed maintenance window.
PURDUE MODEL SEGMENTATION REPORT
====================================
Assessment Date: YYYY-MM-DD
Facility: [Plant Name]
CURRENT STATE:
Network Type: [Flat/Partially segmented/Fully segmented]
IT-OT Boundary: [None/Firewall/DMZ with dual firewall]
Direct IT-to-PLC paths: [count]
RECOMMENDED ARCHITECTURE:
Level 0-1: VLAN 110 (Control Network)
Level 2: VLAN 120 (Supervisory Network)
Level 3: VLAN 130 (Operations Network)
Level 3.5: VLAN 150 (IT/OT DMZ)
Level 4-5: VLAN 200+ (Enterprise)
DMZ COMPONENTS:
- Historian Replica Server
- Jump Server (MFA-enabled)
- Patch Staging Server
- AV Relay Server
FIREWALL RULES: [count] rules generated
MIGRATION STEPS: [count] phases planned
tools
Conduct comprehensive GDPR compliance assessments by evaluating data processing activities against EU Regulation 2016/679, including Article 30 records of processing, lawful basis validation, data subject rights implementation, Data Protection Impact Assessments (DPIAs) under Article 35, breach notification procedures, international transfer safeguards (SCCs, adequacy decisions), and technical/organizational measures under Article 32. Use when processing personal data of EU residents, preparing for supervisory authority audits, implementing privacy-by-design for new systems, scoping compliance gaps for M&A due diligence, assessing third-party processors, or responding to data subject access requests at scale. Incorporates 2026 guidance from ICO, EDPB, and post-Data (Use and Access) Act 2025 UK-GDPR considerations. Do not use for implementing specific Article 32 controls — use implementing-gdpr-data-protection-controls; or for DSAR automation — use implementing-gdpr-data-subject-access-request.
tools
Parse Windows forensic artifacts—$MFT/$J (MFTECmd), Prefetch (PECmd), registry hives (RECmd), shellbags, and Amcache—into normalized CSV/JSON with Eric Zimmerman's EZ Tools, then load results into Timeline Explorer for analysis. Use during DFIR/incident-response investigations, after triage collection (e.g. with KAPE), to establish program execution, file/folder access, and persistence evidence from acquired forensic images.
development
Build automated multi-turn adversarial attacks against conversational LLM targets using Microsoft PyRIT's RedTeamingOrchestrator, CrescendoOrchestrator (gradual escalation), and TreeOfAttacksWithPruningOrchestrator (adaptive branching), with scorer feedback loops and persisted conversation memory. Use when single-shot LLM scanning is insufficient and you need multi-turn, scorer-driven AI red-team campaigns against a chatbot or agent.
testing
Stand up MISP, enable and cache curated threat feeds (CIRCL, abuse.ch, Feodo Tracker), apply warninglists to suppress false positives, query indicators with PyMISP, and export attributes as auto-generated Suricata/Sigma/Wazuh detection rules. Use when maturing a MISP instance to actively drive detection, curating threat feeds with quality controls, or automating IOC-to-detection pipelines for the SIEM/IDS.