| name | Social Alpha Miner |
| description | Automated social media monitor that mines trading strategies from MoltBook and Colosseum user comments. |
Social Alpha Miner 🕵️♂️
This skill provides the capability to monitor social platforms (MoltBook and Colosseum), analyze the discourse using NLP (Natural Language Processing), and extract beneficial trading strategies or market sentiment.
Architecture
- Miner (
social_miner.mjs): The main process that polls APIs.
- Analyzer (
social_sentiment.js): The pipeline module that classifies text and extracts instructions.
- Knowledge Base (
alpha_db.json): A persistent store of extracted wisdom.
Platforms Supported
- Twitter / X VIPs: Analysis of high-impact profiles (e.g., Donald Trump, Michael Saylor, WatcherGuru).
- Colosseum: Agent-to-Agent forum monitoring.
- MoltBook: Social sentiment from the Sidex ecosystem.
Usage
Run the miner script directly:
node skills/social_alpha_miner/scripts/social_miner.mjs
Configuration
Ensure you have the following environment variables set in your .env or .env.hackathon file:
MOLTBOOK_API_KEY: For accessing MoltBook API (if private).COLOSSEUM_API_KEY: For accessing Colosseum Agent API.
Features
- Multi-Platform Support: Extensible class structure for adding more sources (X/Twitter, Discord).
- Sentiment Analysis: Filters noise and saves only high-confidence trading alpha.
- Impact Engine:
- Automatically classifies news as
CRITICAL, HIGH, or MEDIUM.
- Detects market direction (
BULLISH vs BEARISH).
- Gives higher weight to VIP accounts.
- Instruction Storage: Saves standardized instructions (e.g.,
URGENT_BULLISH_ACTION, CONSIDER_SHORT) for the trading bot to consume.
