Click on "Install 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., "@Xcelium MCP Server find when top.dut.error_flag becomes 1'b1 using binary search"
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.
MCP (Model Context Protocol) server that enables AI assistants to control Cadence Xcelium/SimVision in real time via a Tcl socket bridge. Supports automated RTL/gate-level debugging with watchpoints, binary search, checkpoints, and SHM probe control.
┌──────────────┐ stdio ┌───────────────────┐ TCP ┌─────────────────────┐
│ AI Assistant │ <------> │ Python FastMCP │ <-----> │ mcp_bridge.tcl │
│ (Claude, etc) │ │ Server (25 tools) │ :9876 │ inside xmsim/SV │
└──────────────┘ └────────────────────┘ │ (13 meta commands) │
└─────────────────────┘Related MCP server: Verilator MCP Server
pip install -e .
# With screenshot support (requires ghostscript)
pip install -e ".[screenshot]"
# With dev dependencies
pip install -e ".[dev]"Load the Tcl bridge when launching xmsim or SimVision:
# Batch mode (no GUI license needed)
xmsim -64bit -input mcp_bridge.tcl top
# SimVision GUI mode
simvision -64bit -input mcp_bridge.tcl dump.shmThe bridge listens on TCP port 9876 by default. Override with:
export MCP_BRIDGE_PORT=9877Bridge signals readiness by creating /tmp/mcp_bridge_ready_9876.
Claude Code (~/.claude.json):
{
"mcpServers": {
"xcelium-mcp": {
"type": "stdio",
"command": "ssh",
"args": ["-o", "BatchMode=yes", "sim-server", "/path/to/xcelium-mcp"]
}
}
}Claude Desktop (claude_desktop_config.json):
{
"mcpServers": {
"xcelium": {
"command": "xcelium-mcp"
}
}
}# Forward port 9876 in ~/.ssh/config
Host sim-server
LocalForward 9876 localhost:9876Tool | Description |
| Connect to bridge (host, port, timeout) |
| Disconnect (for reconnection only) |
| Run simulation with duration and timeout (default 600s for gate sim) |
| Stop a running simulation |
| Restart from time 0 |
| Get current time, scope, state |
| Set conditional breakpoint |
Tool | Description |
| Read current signal values |
| Get signal type, width, direction |
| Find all drivers (X/Z debugging) |
| List signals in a scope |
| Force a value onto a signal |
| Release a deposited signal |
Tool | Description |
| Add signals to waveform viewer |
| Set waveform time range |
| Set waveform cursor position |
Tool | Description |
| Capture waveform as PNG |
| Full debug snapshot with checklist |
Tool | Description |
| Safe shutdown preserving SHM waveform data |
| Set watchpoint to stop at exact clock edge when condition is true |
| Clear watchpoints (specific ID or all) |
| Enable/disable SHM recording, optionally per scope |
| Save simulation state for later restoration |
| Restore to a saved checkpoint |
| Binary search to find when a condition first becomes true |
connect_simulator()
watch_signal(signal="top.dut.r_state", op="==", value="4'hF")
sim_run(duration="100us") # stops at exact clock edge
get_signal_value(signals=["top.dut.r_state", "top.dut.r_data"])
watch_clear()
shutdown_simulator() # always use this, never disconnectconnect_simulator()
bisect_signal(
signal="top.dut.r_error", op="==", value="1'b1",
start_ns=0, end_ns=1000000, # 0-1ms range
precision_ns=100 # 100ns precision
)
# Returns iteration log + final narrowed time range
shutdown_simulator()connect_simulator()
probe_control(mode="disable") # no SHM recording
sim_run(duration="50ms") # skip uninteresting region
probe_control(mode="enable") # start recording
sim_run(duration="10ms") # capture region of interest
shutdown_simulator()connect_simulator()
sim_run(duration="10ms")
save_checkpoint(name="before_bug")
sim_run(duration="5ms") # analyze bug region
restore_checkpoint(name="before_bug") # go back
sim_run(duration="5ms") # try different analysis
shutdown_simulator()Always specify duration in sim_run to prevent hang on infinite loops
Always use shutdown_simulator to end sessions (preserves SHM data)
Never use disconnect_simulator to end sessions (SHM not flushed)
Gate-level sim: increase timeout with sim_run(timeout=1800) if needed
Bridge ready: check /tmp/mcp_bridge_ready_<port> file instead of TCP ping
The bridge (mcp_bridge.tcl) accepts these meta commands over TCP:
Command | Description |
| Health check |
| Close connection |
| Capture waveform to PostScript |
| Safe shutdown (database close + finish) |
| Non-blocking sim run |
| Query sim time during async run |
| Set signal watchpoint |
| List active watchpoints |
| Delete watchpoints |
| Toggle SHM recording |
| Save checkpoint |
| Restore checkpoint |
| Binary search |
Any other input is evaluated as a raw Tcl/SimVision command.
pytest tests/Python >= 3.10
mcp >= 1.0.0
xmsim or SimVision (Cadence Xcelium) with Tcl console
ghostscript (optional, for EPS to PNG screenshot conversion)
MIT
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
33 tools that make AI write, implement, and verify intent against explicit, testable constraints.
Debug, build, and manage Power Automate cloud flows with AI agents
Build, validate, and deploy multi-agent AI solutions from any AI environment.
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/hslee-cmyk/xcelium-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server