Stop Emotional Trading – Let This AI Handle It Instead

If you’ve ever bought high and sold low, you’re not alone. Emotional trading is the silent killer of portfolios. But what if you could hand over the decision-making to an AI that doesn’t get scared, greedy, or bored?

Meet Honeclaw – an open-source trading assistant that uses machine learning and sentiment analysis to help you make more rational, data-driven trades. It’s not a full autopilot, but it’s a smart co-pilot.

What It Does

Honeclaw is a Python-based tool that scrapes financial news, social media sentiment, and market data to generate trade signals. It combines natural language processing (NLP) with a lightweight ML model to score whether a stock or crypto is likely to move up, down, or sideways.

Instead of you staring at charts all day, Honeclaw does the boring part – scanning thousands of headlines and posts – and gives you a simple “buy/hold/sell” recommendation based on the emotional and factual noise.

Why It’s Cool

• Emotion filter – It explicitly ignores hype and panic. The model is trained to detect when fear or greed is driving the narrative, and it downgrades those signals.
• No black box – The code is fully transparent. You can see exactly how the sentiment score is calculated and which news sources are used. No “trust me bro” ML.
• Plug-and-play for devs – It’s a Python CLI with a simple JSON config. You can run it locally, hook it into your own trading bot, or just use it as a dashboard to double-check your gut feelings.
• Real-time feed – It listens to Twitter, Reddit (top finance subs), and major news APIs. You can add custom sources too.
• Backtesting built in – You can run it on historical data to see how it would have performed before trusting it with real money.

How to Try It

Clone the repo and install the dependencies. No Docker needed, just Python 3.9+ and a few pip install commands.

git clone https://github.com/B-M-Capital-Research/honeclaw
cd honeclaw
pip install -r requirements.txt
python honeclaw.py --config config.json

You’ll need free API keys for NewsAPI and Reddit (optional for Reddit). The config.json has placeholder values – fill them in and you’re ready to run your first sentiment scan. The output is straightforward: a list of assets with a score from -1 to 1, plus a suggested action.

Check the README for a full setup guide.

Final Thoughts

Honeclaw won’t make you a millionaire overnight. But if you’re a developer who trades, it’s a solid tool to add to your stack. Use it as a sanity check before pulling the trigger on a trade. Combine it with your own analysis. It’s free, it’s open source, and it’s built by people who clearly understand that the market is 50% psychology and 50% data.

The code is clean, well-commented, and easy to extend. I might even fork it to add crypto futures signals later. Give it a spin, and maybe finally stop checking your portfolio every 10 minutes.

Found this interesting? Follow us at @githubprojects for more developer tools and open source projects.