Packet Tracer MCP
Provides programmatic control over Cisco Packet Tracer, enabling AI agents to plan, validate, generate, and deploy network topologies; configure routing, switching, DHCP, VLANs, security, IPv6, wireless, and ACL/NAT; inspect live devices, ports, VLANs, and connectivity; trace packets; and save/open real .pkt projects.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Packet Tracer MCPcreate a 3-router OSPF topology with DHCP and deploy it to Packet Tracer"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Install on a fresh copy of ubuntu
1)Install MCP packet tracer
git clone https://github.com/hanley/MCP-PT.git
cd MCP-PT
sudo apt install python3.12-venv
python3 -m venv mcp-pkt
source mcp-pkt/bin/activate
pip install -e .
Related MCP server: Cisco SD-WAN MCP Server
2)Install claude
sudo apt update
sudo apt install curl ca-certificates gnupg -y
Update Claude repository
tmp_key="$(mktemp)"
tmp_gnupg="$(mktemp -d)"
status_code=1
if curl -fsSLo "$tmp_key" https://downloads.claude.ai/claude-desktop/key.asc && \
key_metadata="$(GNUPGHOME="$tmp_gnupg" gpg --batch --show-keys --with-colons "$tmp_key")"; then
fingerprint="$(printf '%s\n' "$key_metadata" | awk -F: '$1 == "fpr" {print $10; exit}')"
printf 'Anthropic key fingerprint: %s\n' "$fingerprint"
if [ "$fingerprint" = "31DDDE24DDFAB679F42D7BD2BAA929FF1A7ECACE" ]; then
sudo install -Dm644 "$tmp_key" /usr/share/keyrings/claude-desktop-archive-keyring.asc
status_code=$?
else
echo "Unexpected Anthropic signing key fingerprint" >&2
fi
else
echo "Anthropic key download or inspection failed" >&2
fi
rm -rf "$tmp_key" "$tmp_gnupg"
[ "$status_code" -eq 0 ]Sign the app
arch="$(dpkg --print-architecture)"
case "$arch" in
amd64|arm64)
echo 'deb [arch=amd64,arm64 signed-by=/usr/share/keyrings/claude-desktop-archive-keyring.asc] https://downloads.claude.ai/claude-desktop/apt/stable stable main' \
| sudo tee /etc/apt/sources.list.d/claude-desktop.list > /dev/null
;;
*)
printf 'Unsupported Claude Desktop architecture: %s\n' "$arch" >&2
false
;;
esac
sudo apt update
apt-cache policy claude-desktop
sudo apt install claude-desktop -y
claude-desktop
3)Install ollama
cd ~
curl -fsSL https://ollama.com/install.sh | sh
export ANTHROPIC_BASE_URL=http://localhost:11434
ollama pull qwen2.5:3b
sudo systemctl edit ollama
[Service]
Environment="OLLAMA_HOST=0.0.0.0:11434"sudo systemctl daemon-reload
sudo systemctl restart ollama
curl http://VM_IP_ADDRESS:11434/api/tags
4)Install mcp ollama
cd ~
sudo apt install npm
git clone https://github.com/MikeyBeez/mcp-ollama.git
cd mcp-ollama
npm install --save-dev @types/node-fetch@2
npm run build
npm uninstall node-fetch
npm install node-fetch@2
rm -rf dist
npm run build
Install mcp packet tracer - continue
cd ~/MCP-PT
source mcp-pkt/bin/activate
pip install -e . => install using local python file
python -m packet_tracer_mcp --stdio
**if working, no error message. Control-C to exit
5)Install Packet Tracer
Download from Cisco Netacad
cd ~/Downloads
sudo apt install ./CiscoPacketTracer_901_Ubuntu_64bit.deb
If import given OVA, OVA comes with all the above installed
Start claude-desktop and login for claude to generate the claude_desktop_config.json.
Shutdown claude-desktop before doing the steps below.
cd MCP-PT
source mcp-pkt/bin/activate
nano /home/analyst/.config/Claude/claude_desktop_config.json
"mcpServers": {
"ollama": {
"command": "node",
"args": ["/home/analyst/mcp-ollama/dist/index.js"],
"env": {
"OLLAMA_BASE_URL": "http://< VM_IP_ADDRESS>:11434"
}
}
},**When using claude-desktop, no way to choose the model, must specify in the prompt such as
"Use the Ollama MCP server to list the available Ollama models.
Use qwen2.5:3b through Ollama to answer my next networking question."
or
"Use the Ollama MCP server and ask qwen2.5:3b:
Explain how to configure a VLAN on a Cisco IOS switch."
nano /home/analyst/.config/Claude/claude_desktop_config.json
,
"packet-tracer": {
"command": "/home/analyst/MCP-PT/mcp-pkt/bin/python",
"args": [
"-m",
"packet_tracer_mcp",
"--stdio"
]
}Use web browser to go to https://github.com/Mats2208/MCP-Packet-Tracer to download V5.2.pts
Add to Scripting > Configure PT Script Modules
Create token on terminal if No MCP token found on this machine.
Do this on the Ubuntu machine where you run packet_tracer_mcp:
cd MCP-PT
source mcp-pkt/bin/activate
echo "$PT_MCP_BRIDGE_TOKEN"
If it is empty, generate one:
export PT_MCP_BRIDGE_TOKEN=$(python3 -c "import secrets; print(secrets.token_urlsafe(32))")
Need to ask the prompt
"Use the Packet Tracer MCP tool pt_bridge_status and tell me the connection status."
To allow the connection from claude desktop
How to start the system
cd MCP-PT
source mcp-pkt/bin/activate
export PT_MCP_BRIDGE_TOKEN=$(python3 -c "import secrets; print(secrets.token_urlsafe(32))")
python -m packet_tracer_mcp --stdio
open another terminal (don’t start claude from menu)
claude-desktop
import packet tracer skill into claude
start packet tracer
if MCP control center still not connected,
send a prompt in claude-desktop
“Use the ollama MCP and Packet Tracer MCP tool pt_bridge_status and tell me the connection status.”
Restart packet
Example Prompt:
Using ollama mcp, Using the Packet Tracer MCP tools, create a topology plan with
2 routers, 2 switches and 4 PCs.
Do not deploy it yet.
Using ollama mcp, Validate the topology.
Using ollama mcp, deploy the topology into Packet Tracer.
This server cannot be deployed
Maintenance
Related MCP Connectors
AI helper for choosing and speccing network & security products from Cisco, Arista, Juniper and more
Network hardware BOMs, built inside your AI assistant
Provides capabilities that let LLM agents perform a range of infrastructure management tasks.
Agent-readiness scanner (0-5 score), robots.txt + llms.txt generators, managed agent enablement.
Related MCP Servers
- AlicenseNot gradedqualityAmaintenanceEnables AI assistants to interact with Cisco Modeling Labs (CML) using natural language, allowing creation, management, and automation of network labs.423 PyPI75BSD 2-Clause "Simplified"
- AlicenseBqualityDmaintenanceEnables natural language automation of Cisco SD-WAN vManage, including device management, template deployment, policy configuration, monitoring, and software upgrades.2MIT
- FlicenseNot gradedqualityDmaintenanceEnables AI models to generate, validate, and deploy network topologies directly into Cisco Packet Tracer from natural-language prompts.-
- AlicenseNot gradedqualityCmaintenanceEnables AI agents to directly inspect, configure, test, and orchestrate network devices through pyATS/Genie, including show command parsing, config rollback, feature state learning/diffing, fleet-wide command fan-out, and declarative test execution.MIT