Quick Start¶

Once your MCP client is connected, just talk to your AI.

Example prompt¶

"Build a network with 2 routers, 2 switches, 4 PCs, DHCP and static routing."

The LLM calls pt_full_build, which runs the whole pipeline — plan → validate → generate → explain → estimate — and returns a complete summary:

Devices (8):  R1, R2 (2911), SW1, SW2 (2960-24TT), PC1..PC4 (PC-PT)
Links   (7):  R1<->R2 (cross), R1<->SW1 (straight), R2<->SW2 (straight),
              SW1<->PC1, SW1<->PC2, SW2<->PC3, SW2<->PC4 (straight)

IP Plan:
  LAN 1:  192.168.0.0/24 -- R1 Gig0/1: .1, PC1: .2, PC2: .3
  LAN 2:  192.168.1.0/24 -- R2 Gig0/1: .1, PC3: .2, PC4: .3
  Link:   10.0.0.0/30    -- R1 Gig0/0: .1, R2 Gig0/0: .2

DHCP:   Pools on R1 and R2
Routes: Bidirectional static routes

The recommended flow¶

For a new topology:

pt_list_devices → pt_plan_topology → pt_validate_plan → pt_live_deploy

…or the one-shot shortcut:

pt_full_build        # plan + validate + generate + (optional) deploy

To edit a topology already open in PT:

pt_bridge_status → pt_query_topology → pt_add_device / pt_add_link /
                                        pt_rename_device / pt_move_device /
                                        pt_delete_link / pt_delete_device

Always discover before you build

The server instructs the LLM to call pt_list_devices (and pt_list_modules for expansion cards) first, so it uses real model names, ports and cables from the catalog instead of guessing. See the Tool reference.

Three ways to get a topology into Packet Tracer¶

Live deploy (pt_live_deploy) — streams commands straight into a running PT via the bridge. See Live Deploy Setup. (Best experience.)
Clipboard (pt_full_build / pt_deploy) — copies a PTBuilder script to your clipboard; paste it into PT's Builder Code Editor and click Run.
Export to disk (pt_export) — writes the PTBuilder script, per-device IOS configs and the plan JSON to projects/<name>/ for reuse.

Ready for real-time? Continue to Live Deploy Setup.