| name | cisco-network-diagram |
| description | Generate Draw.io network topology diagrams from Cisco router and switch configurations. Use when asked to visualize network topology, create network diagrams, map network infrastructure, or generate visual representations from Cisco device configs. Supports parsing show commands, configuration files, CDP/LLDP neighbor data, and routing protocol information to automatically create professional network diagrams. |
| license | MIT |
Cisco Network Diagram Generator
Generate professional Draw.io network topology diagrams from Cisco device configurations using the drawio-network-plot Python library.
Overview
This skill automates the creation of network topology diagrams by:
- Parsing Cisco device configurations to extract topology data
- Identifying devices, interfaces, connections, and network relationships
- Generating Draw.io XML diagrams with proper device icons and layouts
- Supporting multi-layer visualization (physical, logical, L2, L3)
Assumption: Device configurations have already been collected. This skill focuses on parsing and visualization, not SSH access or data collection.
Installation
Install required library:
pip install drawio-network-plot --break-system-packages
Quick Start Workflow
- Parse configurations - Extract topology data from config files
- Build topology model - Create device and connection objects
- Generate diagram - Use drawio-network-plot to create XML
- Save output - Write to
/mnt/user-data/outputs/network-diagram.drawio
Parsing Cisco Configurations
Key Data to Extract
Extract these elements from Cisco configs:
Device Information:
- Hostname:
hostname <name> - Device type: Infer from model (router/switch/firewall/L3 switch)
- Management IP:
interface <mgmt>+ip address
Interface Details:
- Interface name, status, IP address, subnet mask
- VLANs:
switchport access vlan,switchport trunk allowed vlan - Descriptions:
description <text> - Speed/duplex settings
Layer 2 Connections:
- CDP neighbors: Parse
show cdp neighbors detailoutput - LLDP neighbors: Parse
show lldp neighbors detailoutput - Port channels:
interface Port-channel, member interfaces
Layer 3 Information:
- Static routes:
ip route - OSPF:
router ospf, network statements, areas - BGP:
router bgp, neighbors, AS numbers - EIGRP:
router eigrp, network statements - HSRP/VRRP:
standby,vrrpconfigurations
Important: Use references/cisco-parser.md for detailed parsing patterns and regular expressions for each configuration type.
Using drawio-network-plot
Basic Structure
from drawio_network_plot import create_graph, Node, Edge
# Create graph
graph = create_graph(filename="network.drawio", direction="LR")
# Add devices as nodes
router = Node(
id="rtr1",
label="Core-Router-01",
device_type="router",
ip_address="10.0.0.1"
)
graph.add_node(router)
# Add connections as edges
link = Edge(
source="rtr1",
target="sw1",
label="Gi0/0/0 → Gi1/0/1\n10.1.1.0/30"
)
graph.add_edge(link)
# Generate diagram
graph.draw()
Device Types and Icons
Map Cisco devices to appropriate icons:
- Router:
device_type="router"- Routers, multilayer switches doing L3 - Switch:
device_type="switch"- L2 switches, access switches - Firewall:
device_type="firewall"- ASA, Firepower - Server:
device_type="server"- Services, appliances - Cloud:
device_type="cloud"- Internet, WAN, external networks - PC:
device_type="pc"- End devices, workstations
Layout Strategies
Hierarchical (Top-Down):
graph = create_graph(filename="network.drawio", direction="TB")
Best for: Core → Distribution → Access topologies
Left-to-Right:
graph = create_graph(filename="network.drawio", direction="LR")
Best for: Service chains, data flows, sequential processing
Manual Positioning:
node = Node(id="sw1", label="Switch", x=100, y=200, width=120, height=80)
Best for: Complex topologies, specific physical layouts, data center rows
Edge Labeling Best Practices
Include relevant information on connection labels:
# L3 Point-to-Point
Edge(label="Gi0/0 → Gi0/1\n10.1.1.0/30\nOSPF Area 0")
# L2 Trunk
Edge(label="Trunk\nVLANs: 10,20,30,100\n1Gbps")
# Port Channel
Edge(label="Po1 (LACP)\nGi0/1-2\n2Gbps")
Complete Workflow Example
Use scripts/generate_network_diagram.py for a complete implementation:
python scripts/generate_network_diagram.py <config_directory> <output_file>
Or follow this manual workflow:
import re
from drawio_network_plot import create_graph, Node, Edge
# 1. Parse configurations
devices = {}
connections = []
for config_file in config_files:
# Parse device info
hostname = extract_hostname(config_file)
device_type = infer_device_type(config_file)
mgmt_ip = extract_mgmt_ip(config_file)
devices[hostname] = {
'type': device_type,
'mgmt_ip': mgmt_ip,
'interfaces': extract_interfaces(config_file)
}
# Parse CDP neighbors to find connections
cdp_neighbors = parse_cdp_neighbors(config_file)
for neighbor in cdp_neighbors:
connections.append({
'source': hostname,
'source_int': neighbor['local_interface'],
'target': neighbor['device_id'],
'target_int': neighbor['remote_interface']
})
# 2. Create graph
graph = create_graph(filename="network.drawio", direction="TB")
# 3. Add devices
for hostname, info in devices.items():
node = Node(
id=hostname,
label=f"{hostname}\n{info['mgmt_ip']}",
device_type=info['type']
)
graph.add_node(node)
# 4. Add connections (deduplicate bidirectional links)
seen_connections = set()
for conn in connections:
conn_key = tuple(sorted([
f"{conn['source']}:{conn['source_int']}",
f"{conn['target']}:{conn['target_int']}"
]))
if conn_key not in seen_connections:
edge = Edge(
source=conn['source'],
target=conn['target'],
label=f"{conn['source_int']} ↔ {conn['target_int']}"
)
graph.add_edge(edge)
seen_connections.add(conn_key)
# 5. Generate diagram
graph.draw()
Advanced Features
Multi-Layer Diagrams
Create separate diagrams for different network layers:
Physical Layer: All devices and physical connections Logical Layer: VLANs, broadcast domains, subnets Layer 3: Routers, routing protocols, subnets Security Zones: Firewalls, ACLs, security boundaries
Grouping and Containers
Use containers for logical grouping:
from drawio_network_plot import Container
# Create data center container
dc1 = Container(
id="dc1",
label="Data Center 1",
children=["rtr1", "sw1", "sw2", "fw1"]
)
graph.add_container(dc1)
Styling and Customization
Customize colors for different device roles:
node = Node(
id="core-rtr",
label="Core Router",
device_type="router",
fill_color="#FF6B6B", # Red for core devices
font_color="#FFFFFF"
)
Adding Metadata
Include additional information as node properties:
node = Node(
id="rtr1",
label="Core-Router-01",
device_type="router",
metadata={
"model": "ISR4451-X",
"ios_version": "17.6.3",
"location": "Building A - MDF",
"serial": "FDO2148B0X1"
}
)
Handling Complex Topologies
Large Networks (50+ devices)
For large topologies:
- Create hierarchical sub-diagrams by location or function
- Use containers to group related devices
- Generate separate diagrams for L2 and L3 views
- Create summary diagrams showing only core infrastructure
Redundant Paths
Show redundancy clearly:
# Primary path
edge1 = Edge(source="rtr1", target="rtr2", label="Primary\nGi0/0", style="solid")
# Backup path
edge2 = Edge(source="rtr1", target="rtr2", label="Backup\nGi0/1", style="dashed")
MPLS/VPN Topologies
Represent virtual connections:
# VPN tunnel
vpn_edge = Edge(
source="site1-rtr",
target="site2-rtr",
label="IPsec VPN\n10.100.0.0/24",
style="dashed",
color="#0066CC"
)
Output and Delivery
- Save to outputs directory:
import shutil
shutil.move("network.drawio", "/mnt/user-data/outputs/network-diagram.drawio")
Generate multiple formats if needed:
- Primary:
.drawio(editable XML) - Export via Draw.io: PNG, SVG, PDF
- Primary:
Provide user with link:
print("[View your network diagram](computer:///mnt/user-data/outputs/network-diagram.drawio)")
Troubleshooting
Issue: Overlapping nodes Solution: Use manual positioning or adjust layout direction
Issue: Missing connections
Solution: Verify CDP/LLDP data parsing, check for hostname mismatches
Issue: Incorrect device types Solution: Review device type inference logic, use explicit type mapping
Issue: Too much information on diagram Solution: Create multiple focused diagrams (L2, L3, security) instead of one complex diagram
Best Practices
- Start simple: Begin with physical connectivity, add layers incrementally
- Validate data: Cross-reference parsed data with source configs before generating diagram
- Use meaningful labels: Include interface names, IP addresses, and protocol info
- Group logically: Use containers for sites, VLANs, or security zones
- Handle missing data gracefully: Use placeholder values when CDP/LLDP is unavailable
- Document assumptions: Note which connections are discovered vs. inferred
- Provide context: Add title, legend, and timestamp to diagrams
Reference Files
references/cisco-parser.md: Detailed regex patterns and parsing functions for all Cisco configuration typesreferences/drawio-examples.md: Complete working examples for different network scenariosscripts/generate_network_diagram.py: Production-ready script with full parsing and generation logic
Common Use Cases
Use Case 1: Data Center Topology
# Create top-down diagram with core, aggregation, and access layers
# Group servers by rack, show redundant uplinks
Use Case 2: Branch Office WAN
# Show hub-and-spoke topology with MPLS/VPN connections
# Highlight primary and backup WAN links
Use Case 3: VLAN Visualization
# Create L2 diagram showing VLAN trunks and access ports
# Color-code by VLAN or security zone
Use Case 4: Routing Protocol View
# Show OSPF areas, BGP peerings, route reflectors
# Label edges with metrics and next-hops
Next Steps After Generation
- Open diagram in Draw.io (web: app.diagrams.net)
- Manually refine layout and positioning as needed
- Add annotations, labels, or documentation
- Export to PNG/PDF for documentation or presentations
- Version control the .drawio file for change tracking