Amateur radio operators — hams — have been building their own digital infrastructure since before the internet existed. Packet radio in the 1980s. APRS position reporting in the 1990s. Digital modes like FT8 and JS8Call in the 2010s. Now, MCP.
qso-graph (io.github.qso-graph) registered ten MCP servers in a single batch — each one bridging AI agents to a different amateur radio data source. Together they form the most complete niche suite ever registered on the MCP registry.
The Ten Servers
| Server | Score | What It Does |
|---|---|---|
| adif-mcp | 61 | ADIF 3.1.6 parser and validator for amateur radio QSO logs. Full spec coverage. |
| eqsl-mcp | 54 | eQSL.cc integration. Inbox download, QSO verification, Authenticity Guaranteed status. |
| hamqth-mcp | 54 | HamQTH.com callsign lookup, DX spots, Reverse Beacon Network data. |
| iota-mcp | 52 | Islands on the Air. Group lookup, island search, DXCC mapping, nearby groups. |
| lotw-mcp | 54 | ARRL Logbook of The World. QSO query, QSL status, DXCC credit tracking. |
| pota-mcp | 54 | Parks on the Air. Park lookup, activator and hunter stats, live spots. |
| qrz-mcp | 54 | QRZ.com callsign lookup, DXCC resolution, logbook queries via QRZ XML API. |
| solar-mcp | 54 | Solar indices and space weather. SFI, SSN, Kp, DSCOVR, NOAA alerts. |
| sota-mcp | 54 | Summits on the Air. Summit lookup, activator stats, chaser records, spots. |
| wspr-mcp | 54 | WSPR beacon analytics. Band openings, path analysis, solar correlation. |
All ten are Python, all on PyPI, all stdio transport. No flags on any of them.
Why This Matters
If you're not a ham, most of those acronyms mean nothing. That's the point. This is tooling built for a community of roughly 3 million licensed amateur radio operators worldwide — people who communicate by bouncing radio signals off the ionosphere, who track contacts with stations in every country, who hike to mountaintops and set up antennas in national parks to make contacts for points.
The data sources these servers bridge are deeply specialized:
- ADIF (Amateur Data Interchange Format) is the standard log format for recording radio contacts. Every ham logs every contact — callsign, frequency, mode, signal report, time, location. The adif-mcp server parses and validates these logs to the 3.1.6 spec.
- LOTW (Logbook of The World) is the ARRL's QSO confirmation system. Two stations both upload their logs, and if the details match, the contact is confirmed — which counts toward awards like DXCC (contacting 100+ countries).
- POTA and SOTA — Parks on the Air and Summits on the Air — are activity programs where operators activate parks and mountain summits. Live spots show who's currently on the air and on what frequency.
- WSPR (Weak Signal Propagation Reporter) is a network of beacons that continuously transmit low-power signals. By analyzing which beacons are heard where, you can map real-time HF propagation conditions across the globe.
- Solar indices — solar flux, sunspot number, K-index — directly affect radio propagation. Higher solar activity means better long-distance HF communication. Hams check these numbers like traders check the S&P 500.
The Use Case
An AI agent connected to these ten servers can answer questions that currently require visiting six different websites and cross-referencing data manually:
- "What bands are open right now for a contact between California and Japan?"
- "Show me my unconfirmed LOTW contacts from the last month."
- "Who's activating POTA parks near me right now?"
- "Parse my ADIF log and tell me which DXCC entities I'm missing for the award."
- "How does today's solar flux compare to last week, and what does that mean for 20-meter propagation?"
This is not a toy. Ham radio contesting, award chasing, and propagation analysis involve real data analysis across multiple sources. An AI assistant that can query all of these simultaneously is genuinely useful to the community.
Also in RF This Week
The ham radio suite didn't arrive alone. IONIS-AI (io.github.IONIS-AI/ionis-mcp) registered an HF propagation analytics server with 175 million WSPR and Reverse Beacon Network data points, solar correlation analysis, and band opening detection. Score: 54. And rftools (io.github.antonpogrebenko-public/rftools) brought 197 RF and electronics calculators — microstrip design, link budget analysis, filter calculations, power conversions. Score: 57.
Radio frequency engineering is having its MCP moment. Three independent developers, in the same week, decided AI agents should understand the electromagnetic spectrum.
Sources: qso-graph — GitHub · Scorecard: io.github.qso-graph (scores 52–61) · IONIS-AI — Scorecard (score 54) · rftools — Scorecard (score 57)