# AutoHedge Sentiment Agent: Social Media Platforms and Scoring Explained > Discover how the AutoHedge Sentiment Agent analyzes Reddit, Twitter, and StockTwits, and understand its 0-to-1 sentiment scoring system for actionable market insights. - Repository: [Swarms/AutoHedge](https://github.com/The-Swarm-Corporation/AutoHedge) - Tags: deep-dive - Published: 2026-09-08 --- **The AutoHedge Sentiment Agent monitors Reddit, Twitter, and StockTwits to calculate quantitative sentiment scores on a 0-to-1 scale, where 0 represents extremely negative and 1 represents extremely positive sentiment.** The-Swarm-Corporation/AutoHedge is an open-source algorithmic trading framework that deploys specialized AI agents to analyze market data and social sentiment. The Sentiment Agent aggregates retail investor conversations from major financial social platforms to generate normalized market mood indicators that drive automated trading decisions. ## Monitored Social Media Platforms According to the `SENTIMENT_PROMPT` defined in [`autohedge/prompts.py`](https://github.com/The-Swarm-Corporation/AutoHedge/blob/main/autohedge/prompts.py) (lines 47-48), the Sentiment Agent explicitly tracks three social media platforms to capture comprehensive retail investor sentiment. ### Reddit The agent analyzes discussion threads and subreddits related to specific stocks and market trends. Reddit serves as a primary source for grassroots investor sentiment, meme stock movements, and emerging market narratives that often precede mainstream financial coverage. ### Twitter The system processes tweets and hashtags mentioning ticker symbols to capture real-time market reactions and breaking news sentiment. This platform provides immediate insight into broad retail investor mood swings and viral financial discussions. ### StockTwits As a dedicated stock-focused social network, StockTwits offers targeted financial conversations between traders and investors. The agent monitors this platform for professional sentiment analysis and ticker-specific discourse that may not appear on general social networks. ## Sentiment Score Calculation Methodology The Sentiment Agent produces a **quantitative sentiment score** normalized to a **0-to-1 scale** (as specified in [`autohedge/prompts.py`](https://github.com/The-Swarm-Corporation/AutoHedge/blob/main/autohedge/prompts.py) lines 49-50). - **0**: Extremely negative sentiment - **1**: Extremely positive sentiment The score aggregates the **polarity** and **intensity** of collected news articles and social media posts. This quantitative approach converts qualitative social media chatter into a normalized metric representing the overall market mood for specific securities. ## Implementation in the AutoHedge Codebase The Sentiment Agent's behavior is implemented across two critical files that define both the analytical logic and the runtime interface. ### Configuration in prompts.py The platform list and scoring methodology are hardcoded in the system prompt. In [`autohedge/prompts.py`](https://github.com/The-Swarm-Corporation/AutoHedge/blob/main/autohedge/prompts.py), the `SENTIMENT_PROMPT` specifies: - The three monitored platforms (Reddit, Twitter, StockTwits) - The 0-1 scoring range and its interpretation - The aggregation logic for polarity and intensity ### Agent Initialization in workers.py The `sentiment_agent` object is declared in [`autohedge/workers.py`](https://github.com/The-Swarm-Corporation/AutoHedge/blob/main/autohedge/workers.py) with its system prompt and model configuration pre-loaded. This encapsulation allows developers to invoke sentiment analysis without additional prompt engineering. ### Practical Usage Example ```python from autohedge.workers import sentiment_agent # Example input: ticker, recent news & social posts (simplified) message = """\ Ticker: AAPL News: Apple released its latest iPhone with strong sales forecasts. Social: Reddit thread trending bullish, Twitter sentiment mixed, StockTwits praising earnings.""" # Run the agent – it returns a JSON-compatible sentiment breakdown result = sentiment_agent.run(message) print(result) # → overall score, news score, social score, etc. ``` The `sentiment_agent.run()` method processes the input through the configured LLM using the `SENTIMENT_PROMPT`, automatically applying the platform monitoring rules and 0-1 scoring logic defined in the source code. ## Summary - The AutoHedge Sentiment Agent monitors **three platforms**: Reddit, Twitter, and StockTwits. - Sentiment scores are calculated on a **0-to-1 scale**, where 0 is extremely negative and 1 is extremely positive. - The scoring methodology aggregates **polarity and intensity** from both news articles and social media posts. - Configuration resides in [`autohedge/prompts.py`](https://github.com/The-Swarm-Corporation/AutoHedge/blob/main/autohedge/prompts.py) (lines 47-50), while the agent implementation is in [`autohedge/workers.py`](https://github.com/The-Swarm-Corporation/AutoHedge/blob/main/autohedge/workers.py). - The agent is pre-configured and ready to use via the `sentiment_agent.run()` method without additional prompt engineering. ## Frequently Asked Questions ### Which social media platforms does the Sentiment Agent monitor? The Sentiment Agent monitors Reddit, Twitter, and StockTwits. These platforms are explicitly listed in the `SENTIMENT_PROMPT` located in [`autohedge/prompts.py`](https://github.com/The-Swarm-Corporation/AutoHedge/blob/main/autohedge/prompts.py) at lines 47-48, covering both general social networks and specialized financial communities. ### How are sentiment scores calculated in AutoHedge? Sentiment scores are calculated on a quantitative 0-to-1 scale, where 0 represents extremely negative sentiment and 1 represents extremely positive sentiment. The calculation aggregates the polarity and intensity of collected news articles and social media posts, as defined in [`autohedge/prompts.py`](https://github.com/The-Swarm-Corporation/AutoHedge/blob/main/autohedge/prompts.py) lines 49-50. ### Where is the Sentiment Agent configured in the codebase? The Sentiment Agent is configured in [`autohedge/prompts.py`](https://github.com/The-Swarm-Corporation/AutoHedge/blob/main/autohedge/prompts.py), which defines the `SENTIMENT_PROMPT` containing platform specifications and scoring rules. The agent instance itself is created in [`autohedge/workers.py`](https://github.com/The-Swarm-Corporation/AutoHedge/blob/main/autohedge/workers.py) with this prompt and model configuration bound to it. ### What does a score of 0.5 indicate in AutoHedge sentiment analysis? A score of 0.5 indicates neutral sentiment, representing the midpoint between extremely negative (0) and extremely positive (1). This suggests balanced polarity and intensity in the aggregated social media and news data for the analyzed ticker.