AI4Finance-Foundation / AI4Finance-Foundation/FinRobot

Add historical chart pattern context to equity research reports

Aperta
#95 0 commenti 0 reazioni 0 assegnatari Vedi su GitHub
Lingua principale
Jupyter Notebook
Stelle
8k
Fork
1.4k
Metriche di merge delle PR
Nessuna PR unita negli ultimi 30g

Descrizione

Great project — the automated equity research pipeline is impressive, especially the multi-source data aggregation.

I built [Chart Library](https://chartlibrary.io), an API that finds historically similar chart patterns and returns what happened next. It could add a "historical pattern context" section to generated research reports — e.g. "the current NVDA chart most closely resembles 10 prior setups, 8 of which were higher 5 days later."

Adding it as a data source would be straightforward:

```python
import requests
# Returns similar patterns, forward returns, and market regime context
data = requests.get("https://chartlibrary.io/api/v1/intelligence/NVDA").json()
```

There's also an MCP server (`pip install chartlibrary-mcp`) with 19 tools for more granular queries — individual pattern matches, regime analysis, specific timeframes, etc.

Free tier: 200 calls/day, no signup. Covers 15K stocks with 10 years of minute-bar data and 24M pre-computed pattern embeddings.

This is historical pattern data, not predictions — so it fits well in a research report context where you want to present what has happened before and let the reader draw conclusions. Happy to help if you want to explore it.

Guida per i contributori

Nessuna guida per i contributori indicizzata per questo repository

Valutazione

Questa issue non è ancora stata valutata.

Ricevi le nuove issue nella tua casella

Un breve riepilogo di issue GitHub adatte ai principianti.