wspr-mcp
Enables querying WSPR beacon data including live spots, band activity, top beacons, propagation paths, and SNR trends through any MCP-compatible AI assistant.
README
<!-- mcp-name: io.github.qso-graph/wspr-mcp -->
wspr-mcp
MCP server for WSPR (Weak Signal Propagation Reporter) beacon data — live spots, band activity, top beacons, propagation paths, SNR trends, and more through any MCP-compatible AI assistant.
Data from wspr.live (~2.7 billion spots, 2008-present). Part of the qso-graph project. No authentication required — all public data.
Install
pip install wspr-mcp
Tools
| Tool | Description | Key Parameters |
|---|---|---|
wspr_spots |
Recent WSPR spots with flexible filtering | callsign, band, hours, grid, min/max SNR, min distance |
wspr_band_activity |
Per-band spot counts, station counts, distances, and SNR | hours |
wspr_top_beacons |
Top transmitters ranked by spot count or max distance | band, hours, sort_by (spots/distance), limit |
wspr_top_spotters |
Top receivers ranked by spot count or max distance | band, hours, sort_by (spots/distance), limit |
wspr_propagation |
Propagation between two locations — auto-widens to field-level proxy for sparse endpoints | tx, rx, band, hours |
wspr_grid_activity |
All WSPR activity in/out of a Maidenhead grid square | grid (2 or 4 char), band, hours |
wspr_longest_paths |
Longest distance WSPR paths in a time window | band, hours, min_distance, limit |
wspr_snr_trend |
Hourly SNR trend for a specific path over time | tx, rx, band, hours |
get_version_info |
Service version + upstream spec version (fleet identity attestation) | — |
What is WSPR?
WSPR beacons transmit a 2-minute encoded signal at very low power (typically 200 mW to 5 W). Each decoded spot proves a propagation path exists between two locations on a specific band. With thousands of beacons worldwide transmitting 24/7, WSPR provides continuous, automated propagation monitoring across all HF bands.
Good Neighbour Policy
wspr.live is a volunteer-run service that mirrors all wsprnet.org data into a public ClickHouse database. We take our responsibility as a good neighbour seriously:
| Measure | Detail |
|---|---|
| Rate limiting | 3 seconds between requests (20 req/min max) |
| Circuit breaker | Opens after 3 consecutive failures; exponential backoff up to 5 minutes. Prevents hammering a struggling service. |
| Time-bounded queries | Every query filters by time (max 72 hours). No unbounded full-table scans. |
| Band filtering | Queries filter by band whenever the user provides one — this hits wspr.live's indexes efficiently. |
| Column selection | We SELECT only the columns each tool needs (8-10 per query), never SELECT *. |
| Result limits | All queries cap results (200 spots, 50 leaderboard entries). |
| Response caching | 2-10 minute TTL per tool. Identical queries within the window hit local cache with zero network traffic. |
| Request timeout | 20-second timeout — we don't hold connections open on a shared service. |
| User-Agent header | Every request identifies itself as wspr-mcp/{version} so the operators can reach us if needed. |
If wspr.live is down or overloaded, the circuit breaker backs off automatically. We don't retry in a tight loop.
Quick Start
No credentials needed — just install and configure your MCP client.
Configure your MCP client
Claude Desktop
Add to claude_desktop_config.json (~/Library/Application Support/Claude/ on macOS, %APPDATA%\Claude\ on Windows):
{
"mcpServers": {
"wspr": {
"command": "wspr-mcp"
}
}
}
Claude Code
Add to .claude/settings.json:
{
"mcpServers": {
"wspr": {
"command": "wspr-mcp"
}
}
}
ChatGPT Desktop
{
"mcpServers": {
"wspr": {
"command": "wspr-mcp"
}
}
}
Cursor
Add to .cursor/mcp.json (project-level) or ~/.cursor/mcp.json (global):
{
"mcpServers": {
"wspr": {
"command": "wspr-mcp"
}
}
}
VS Code / GitHub Copilot
Add to .vscode/mcp.json in your workspace:
{
"servers": {
"wspr": {
"command": "wspr-mcp"
}
}
}
Gemini CLI
Add to ~/.gemini/settings.json (global) or .gemini/settings.json (project):
{
"mcpServers": {
"wspr": {
"command": "wspr-mcp"
}
}
}
Ask questions
"Show me recent WSPR spots on 20m"
"What bands are open right now?"
"Who are the top WSPR beacons on 20m today?"
"What are the longest WSPR paths in the last 24 hours?"
"Is there propagation between Idaho (DN13) and central Europe (JN48)?"
"Any WSPR paths between DN13 and Bouvet (JD15)?" (auto-widens to DN↔JD if no exact match)
"What's happening in grid square JN48?"
"How's the SNR trending between K9AN and G8JNJ on 20m?"
"Who are the best WSPR receivers sorted by distance?"
Testing Without Network
WSPR_MCP_MOCK=1 wspr-mcp
MCP Inspector
wspr-mcp --transport streamable-http --port 8009
Development
git clone https://github.com/qso-graph/wspr-mcp.git
cd wspr-mcp
pip install -e .
License
GPL-3.0-or-later
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。