aigroup-finnhub-mcp
# Finnhub MCP
Route event-driven, news-led, filing, sentiment, crypto, and FX work to `aigroup-finnhub-mcp`.
## Route
1. Identify the asset class first:
- equities and company events
- crypto
- forex
- market calendars or holidays
2. Choose the narrowest Finnhub capability that fits:
- `finnhub_stock_market_data` for quote, profile, candles, or earnings surprise context
- `finnhub_news_sentiment` for news, article sentiment, or insider sentiment
- `finnhub_sec_filings` for filings and filing sentiment
- `finnhub_stock_estimates` or `finnhub_market_events` for analyst and event context
- `finnhub_stock_ownership` for insider, congress, and institutional data
- `finnhub_technical_analysis` for aggregate signals or indicators
- `finnhub_crypto_data` and `finnhub_forex_data` for non-equity assets
3. Use date ranges deliberately. Finnhub has many time-bounded endpoints, so specify `from` and `to` when freshness matters.
4. Summarize signal quality, not just raw output. Call out whether a finding is based on news flow, filings, analyst revisions, or technicals.
## Common Jobs
- Build a recent-news and sentiment brief for a ticker.
- Check SEC filings, filing sentiment, or filing similarity.
- Review insider activity, institutional ownership, or congress trading.
- Pull crypto or forex quotes and short candle history.
- Compare technical aggregate signals with the latest fundamental or news backdrop.
- Fetch earnings, economic, IPO, or FDA calendar context.
## References
- Read [capabilities.md](./references/capabilities.md) for the exposed Finnhub tool groups and suggested routing.
标签
skill
ai