SkippyMCP
An MCP server for controlling Rigol oscilloscopes from an AI assistant. It translates MCP tool calls into SCPI commands over PyVISA.
README
SkippyMCP
An MCP (Model Context Protocol) server for controlling Rigol oscilloscopes from an AI assistant. SkippyMCP translates MCP tool calls into SCPI commands over PyVISA, so an assistant can configure channels and triggers, arm captures, read measurements, grab screenshots, pull waveform data, and read protocol-decode results.
The name is a nod to SCPI — pronounced "skippy" in the test-and-measurement world.

Live capture: an AI assistant configured the channel and trigger, ran the acquisition, measured it (Vpp 1.02 V, ~1 MHz), and pulled this screenshot — all through SkippyMCP's MCP tools.
- Project name: SkippyMCP
- Executable:
skippy-mcp - Primary target: Rigol MSO5204 (MSO5000 series); a per-series dialect layer keeps other Rigol DSO/MSO families addable.
- Status: validated live against a real MSO5204; serves MCP over HTTP (Streamable HTTP) with optional Bearer-key auth and TLS.
Install
python3 -m venv .venv
.venv/bin/pip install -e ".[dev]"
Run
The server speaks MCP over HTTP (Streamable HTTP) at /mcp:
skippy-mcp --host 192.168.1.50 # plain HTTP on 0.0.0.0:8080
skippy-mcp --resource TCPIP0::scope::5555::SOCKET --port 9000
skippy-mcp --config skippy.json # API key / TLS / address from JSON
| Flag | Default | Effect |
|---|---|---|
--host / --resource |
— | Instrument address (CLI overrides the config file). |
--bind |
0.0.0.0 |
HTTP bind address. |
--port |
8080 |
HTTP port. |
--timeout-ms |
5000 | VISA I/O timeout. |
--no-reset |
reset on | Skip *RST on connect; leave the setup untouched. |
--allow-raw-scpi |
off | Register the scpi_raw escape-hatch tool. |
--config <path> |
none | Optional JSON config (below). |
Config file (--config)
All keys optional. No file → plain HTTP, no auth.
{
"host": "192.168.1.50",
"resource": "TCPIP0::192.168.1.50::5555::SOCKET",
"api_key": "your-bearer-token",
"tls": { "cert": "/path/cert.pem", "key": "/path/key.pem" }
}
api_keyset → requireAuthorization: Bearer <key>on every request.tlsset → serve HTTPS directly (no reverse proxy needed).- Address precedence:
--resource>--host> JSONresource> JSONhost.
The startup banner reports the active mode (TLS / API key) and prints an example
smoke-test curl.
Ready-to-edit examples for each mode live in
examples/json-configuration/:
| File | Mode |
|---|---|
http-apikey.json |
HTTP + Bearer API key |
https-tls.json |
HTTPS/TLS, no auth |
https-tls-apikey.json |
HTTPS/TLS + Bearer API key |
Generating a self-signed TLS certificate
For local/LAN testing you can make a self-signed cert and key. Include the
address you'll connect to in the subjectAltName so clients can verify it:
openssl req -x509 -newkey rsa:2048 -nodes -days 365 \
-keyout key.pem -out cert.pem \
-subj "/CN=localhost" \
-addext "subjectAltName=IP:127.0.0.1,DNS:localhost"
Point tls.cert / tls.key at the resulting files. Clients that don't already
trust the cert can be told to with SSL_CERT_FILE=cert.pem. In Docker, the cert
and key must be readable by the container's non-root user (e.g. chmod 644).
For production, prefer a CA-issued certificate over a self-signed one.
Docker
docker build -t skippy-mcp:latest .
# --network host reaches a link-local / same-LAN instrument; -p publishes the API:
docker run --rm --network host skippy-mcp:latest \
--resource TCPIP0::<scope-ip>::5555::SOCKET --port 8080
The image is pure-Python (pyvisa-py) and runs as a non-root user.
Tools
get_identity, configure_channel, configure_logic, configure_trigger,
capture, measure, screenshot, read_waveform, decode_bus, and
(when --allow-raw-scpi) scpi_raw.
Develop / test
.venv/bin/pytest # full suite, no hardware (uses the simulator)
.venv/bin/mypy # strict type-check
.venv/bin/ruff check
Documentation
| Document | Description |
|---|---|
| Initial design | Overview, architecture, tool surface, compatible models, prior art. |
| Detailed design | Layered architecture, transport interface + simulator, dialect layer, error model, tool schemas. |
| Implementation plan | Phased build plan (hardware-free through Phase 6). |
| Validation summary | v0.1.0 test results + live MSO5204 validation. |
| HTTP transport design | HTTP transport, config file, API-key auth, TLS (v0.2.0). |
License
MIT © 2026 Mark Deazley
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.