gnuradio-mcp-server
Enables natural language control of a live GNU Radio SDR flowgraph, allowing users to tune frequencies, adjust gain, capture IQ samples, analyze spectra, and detect signals through an MCP-compatible client.
README
GNU Radio MCP Server
An MCP (Model Context Protocol) server that bridges an LLM to a live GNU Radio SDR flowgraph. Control your software-defined radio — tune frequencies, adjust gain, capture IQ samples, analyse spectra, and sweep for signals — all through natural language via any MCP-compatible client.
Architecture
┌─────────────┐ MCP (stdio / HTTP) ┌──────────────────┐
│ LLM / │◄──────────────────────────────►│ gnuradio_mcp │
│ Claude │ tool calls & │ MCP Server │
│ Desktop │ responses │ │
└─────────────┘ └──┬──────────┬───┘
│ │
XML-RPC │ │ ZMQ SUB
(control) │ │ (IQ data)
▼ ▼
┌────────────────────┐
│ GNU Radio │
│ Companion │
│ Flowgraph │
│ ┌──────────────┐ │
│ │ XMLRPC Server │ │
│ │ ZMQ PUB Sink │ │
│ └──────────────┘ │
└────────┬───────────┘
│
▼
┌────────────────────┐
│ SDR Hardware │
│ (HackRF / RTL-SDR │
│ / USRP / Sim) │
└────────────────────┘
Quick Start (Simulation — No Hardware)
pip install gnuradio-mcp-server
# Open simulation_mcp.grc in GNU Radio Companion and click Run
# (provides two test tones at +10 kHz and +50 kHz from center)
gnuradio-mcp
Quick Start (HackRF)
pip install gnuradio-mcp-server
# Open hackrf_mcp.grc in GNU Radio Companion and click Run & ensure HackRF is connected
GNURADIO_HARDWARE=hackrf gnuradio-mcp
Claude Desktop Integration
Add the following to your claude_desktop_config.json:
{
"mcpServers": {
"gnuradio_mcp": {
"command": "gnuradio-mcp",
"args": [],
"env": {
"GNURADIO_HARDWARE": "hackrf",
"GNURADIO_GR_HOST": "127.0.0.1",
"GNURADIO_GR_XMLRPC_PORT": "8080",
"GNURADIO_GR_ZMQ_PORT": "5555"
}
}
}
}
For HTTP transport:
{
"mcpServers": {
"gnuradio_mcp": {
"url": "http://127.0.0.1:8000/mcp",
"transport": "streamable-http"
}
}
}
Tools Reference
| # | Tool Name | Description | Read-Only |
|---|---|---|---|
| 1 | gnuradio_flowgraph_start |
Start the GNU Radio flowgraph | No |
| 2 | gnuradio_flowgraph_stop |
Stop the GNU Radio flowgraph | No |
| 3 | gnuradio_flowgraph_status |
Get flowgraph status and current parameters | Yes |
| 4 | gnuradio_set_center_frequency |
Set the SDR center frequency | No |
| 5 | gnuradio_set_sample_rate |
Set the SDR sample rate | No |
| 6 | gnuradio_set_gain |
Set the SDR gain (IF/RF/BB stages) | No |
| 7 | gnuradio_set_variable |
Set an arbitrary flowgraph variable | No |
| 8 | gnuradio_get_variable |
Read an arbitrary flowgraph variable | Yes |
| 9 | gnuradio_list_variables |
List all flowgraph variables | Yes |
| 10 | gnuradio_capture_samples |
Capture IQ samples and return a power summary | Yes |
| 11 | gnuradio_get_spectrum_snapshot |
FFT/PSD analysis with peak detection | Yes |
| 12 | gnuradio_detect_signals |
Sweep a frequency range and detect active signals | Yes |
| 13 | gnuradio_ping |
Health check (XML-RPC + ZMQ connectivity) | Yes |
Environment Variables
All variables use the GNURADIO_ prefix:
| Variable | Default | Description |
|---|---|---|
GNURADIO_GR_HOST |
127.0.0.1 |
GNU Radio host address |
GNURADIO_GR_XMLRPC_PORT |
8080 |
XML-RPC server port |
GNURADIO_GR_ZMQ_PORT |
5555 |
ZMQ PUB sink port |
GNURADIO_GR_XMLRPC_TIMEOUT |
5 |
XML-RPC timeout in seconds |
GNURADIO_GR_ZMQ_TIMEOUT_MS |
3000 |
ZMQ receive timeout in ms |
GNURADIO_HARDWARE |
hackrf |
Hardware type: hackrf/rtlsdr/usrp/simulation |
GNURADIO_FREQ_MIN_HZ |
1000000 |
Minimum tunable frequency (Hz) |
GNURADIO_FREQ_MAX_HZ |
6000000000 |
Maximum tunable frequency (Hz) |
GNURADIO_GAIN_MIN_DB |
0 |
Minimum gain (dB) |
GNURADIO_GAIN_MAX_DB |
62 |
Maximum gain (dB) |
GNURADIO_SAMP_RATE_MIN |
100000 |
Minimum sample rate (sps) |
GNURADIO_SAMP_RATE_MAX |
20000000 |
Maximum sample rate (sps) |
GNURADIO_DEFAULT_SAMP_RATE |
2000000 |
Default sample rate for DSP |
GNURADIO_TRANSPORT |
stdio |
Transport: stdio / streamable_http |
GNURADIO_HTTP_PORT |
8000 |
HTTP transport port |
Edge Cases & Known Limitations
-
Sample rate restart: Some SDR hardware (especially RTL-SDR) requires a flowgraph restart after changing the sample rate. The
gnuradio_set_sample_ratetool includes a warning about this. -
ZMQ dtype requirement: The ZMQ PUB Sink in your flowgraph must output
complex64(8 bytes per sample). Misaligned data will raise aGRHardwareError. -
Frequency clamping: Frequencies outside the configured
[freq_min_hz, freq_max_hz]range are rejected before being sent to hardware. Adjust the env vars for your specific hardware. -
Context overflow warning: Capturing more than 65,536 IQ samples triggers a warning, as the response summary may still be large. Use
gnuradio_get_spectrum_snapshotfor analysed results instead. -
XML-RPC not thread-safe: A new
ServerProxyis created per RPC call to avoid shared-state bugs in async contexts. -
ZMQ socket lifecycle: Each
recv_samplescall opens and closes its own ZMQ SUB socket. This is intentional — the slight overhead avoids shared-socket state issues.
Development
# Install in development mode
pip install -e ".[dev]"
# Run tests
pytest tests/ -v
# Run with HTTP transport
GNURADIO_TRANSPORT=streamable_http python -m gnuradio_mcp.server
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.