Open-source multi-agent app lets LLM bull/bear agents debate stocks before trading
kingabzpro · reddit · 2026-10-24
A Redditor released BetterTradingAgents, a free open-source web app for stock research built on multi-agent LLM workflows (FastAPI + CrewAI). The pipeline: analyst agents gather market, technical, news, forecast and sentiment data; bull and bear agents debate the case; a judge agent scores the debate; a manager agent makes the final BUY/SELL/HOLD call with a confidence score. Decisions and reasoning are logged to memory and later graded against actual price action, so the app shows its hit rate over time.
- Runs locally, works with any OpenAI-compatible API, tuned for cheap models — a full first session on DeepSeek costs about $0.01
- Includes paper trading via Alpaca and a walk-forward backtest harness
- Author seeks feedback on whether the debate+judge structure actually improves decisions and on confidence calibration in agent pipelines
More from coding & agent
- Harness engineering setup for Opus 5.5: model tiering, context hygiene and progress files that persist — Arindam_1729 · 2026-10-24
- Two core Claude Code creators announce departures from Anthropic the same day — DrDatta_AIIMS · 2026-10-24
- Devs struggle to juggle multiple AI agents across projects without losing context — model_maverick · 2026-10-24
- Xiamen University's FlowMAS uses GFlowNet to auto-design multi-agent topologies — jiqizhixin · 2026-10-24
- System One Computer Use goes open source: screen-acting agent at ~$0.002 per task — GabGarrett · 2026-10-24
- Dev builds Telegram bot to watch AI labs' RSS feeds every minute — teortaxesTex · 2026-10-24