Repolode
Submit
cablate

cablate/banini-tracker

Tracker that monitors a Taiwanese stock commentator's Facebook posts, runs AI reverse-indicator analysis, pushes alerts to Telegram, Discord and LINE, and scores its own predictions.

View on GitHub
Stars
383
Forks
55
Open issues
6
Last commit
17 days ago

Repolode describes open-source projects and does not give investment advice. Trading and investing can lose money: assess the risk yourself and follow the laws where you live before using this project.

What is it?

banini-tracker fetches posts from a Facebook stock-commentary page, transcribes attached videos, and uses an LLM to extract the stocks, ETFs and commodities mentioned. It then derives reverse-indicator views (for example, a stop-loss suggests a possible rebound) and chained effects across sectors, and sends the results to Telegram, Discord or LINE.

Each extracted call is recorded as a prediction with a baseline closing price, then tracked over five trading days, noting intraday highs and lows. A prediction counts as successful when the reverse direction is correct and the move exceeds ±1%. New predictions on the same stock supersede older ones. The project includes a public SQLite dataset of 345 de-identified predictions from a 2024–2026 backtest.

It can be deployed on Zeabur or Docker, run locally via npx with a web settings page, or used as a CLI alongside Claude Code. The project is for entertainment reference only and is not investment advice.

Key features

  • 01Fetches Facebook posts and transcribes videos through Groq Whisper.
  • 02Extracts stocks, ETFs and commodities with an LLM and derives reverse-indicator views.
  • 03Pushes alerts to Telegram, Discord and LINE.
  • 04Tracks each prediction for five trading days against intraday highs and lows.
  • 05Provides a web settings page and scheduled runs on weekdays and daily.

Use cases

  • Monitoring a specific Taiwanese stock commentator's posts with automated alerts.
  • Studying how a contrarian signal performed in a historical backtest using the public dataset.
  • Running a personal stock-post monitor with Claude Code as the analysis engine.

Quick start

  1. 1

    Start the always-on service with npx (no Docker needed)

    npx @cablate/banini-tracker serve
  2. 2

    Open http://localhost:3000 and enter your API keys and notification channels in the settings page

Things to know

  • Requires an Apify token and an LLM API key, which may incur paid usage costs.
  • Licensed under AGPL-3.0, so modified or deployed versions must be open-sourced under the same license.
  • Focused on one Taiwanese stock commentator's Facebook posts and documented in Chinese.

In collections

Claude Code tools, skills & plugins

Topics

#stock-analysis#web-scraping#llm#telegram-bot#discord-bot#line-bot#transcription#taiwan-stock#apify#claude-code

FAQ

What is banini-tracker?
Tracker that monitors a Taiwanese stock commentator's Facebook posts, runs AI reverse-indicator analysis, pushes alerts to Telegram, Discord and LINE, and scores its own predictions. banini-tracker fetches posts from a Facebook stock-commentary page, transcribes attached videos, and uses an LLM to extract the stocks, ETFs and commodities mentioned. It then derives reverse-indicator views (for example, a stop-loss suggests a possible rebound) and chained effects across sectors, and sends the results to Telegram, Discord or LINE.
Is banini-tracker free to use? What license does it have?
banini-tracker is open source under the AGPL-3.0 license. Check the license terms before using it in commercial products.
Is banini-tracker still maintained?
Actively maintained. The last commit to banini-tracker was on Oct 22, 2026.
What can I use banini-tracker for?
Monitoring a specific Taiwanese stock commentator's posts with automated alerts. Studying how a contrarian signal performed in a historical backtest using the public dataset. Running a personal stock-post monitor with Claude Code as the analysis engine.
How do I install banini-tracker?
Start the always-on service with npx (no Docker needed): npx @cablate/banini-tracker serve Open http://localhost:3000 and enter your API keys and notification channels in the settings page
Where can I get banini-tracker?
The source code and installation instructions are on GitHub: https://github.com/cablate/banini-tracker

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects