DemonDamon/FinnewsHunter

FinnewsHunter: Multi-agent financial intelligence platform powered by AgenticX. Real-time news analysis, sentiment fusion, and alpha factor mining.

What it solves

FinnewsHunter is an enterprise-grade platform designed to transform raw financial news into actionable investment signals. It solves the problem of information overload by automating the monitoring of multiple financial news sources, performing deep sentiment analysis, and assessing market impact to identify potential investment opportunities and risks.

How it works

Built on the AgenticX framework, the system uses a multi-agent architecture (including roles like NewsAnalyst and Researcher) to monitor real-time news streams from sources such as Sina Finance and National Business Daily. It integrates LLMs from various providers (OpenAI, DeepSeek, Kimi, Zhipu, Bailian) to interpret news and assess sentiment. The platform combines these AI insights with real-market data via akshare for K-line analysis and uses a vector database (Milvus) and knowledge graphs to mine alpha signals for quantitative trading.

Who it’s for

This tool is primarily for quantitative traders, financial analysts, and investors who need real-time, AI-driven interpretation of financial news and its correlation with stock market movements.

Highlights

  • Multi-Agent Collaboration: Uses a debate mechanism and specialized agents to analyze news sentiment and market impact.
  • Real-time Monitoring: Automatically crawls and deduplicates news from 10 major financial sources every minute.
  • Multi-LLM Support: One-click switching between five major LLM providers for analysis.
  • Integrated Market Data: Combines news analysis with real-time stock K-line charts and fuzzy stock search for 5,000+ A-share stocks.
  • Production-Ready Stack: Deploys via Docker Compose with a full stack including FastAPI, PostgreSQL, Milvus, and Redis.

Related

  • Project
    octotools/octotools

    An agentic framework that uses standardized tool cards and a dual-level planner to enable LLMs to perform complex reasoning across diverse domains without additional training.

  • Project
    zhengkid/Dream-RSI

    Dream-RSI is a framework for recursive self-improvement that uses discovery history as a replay simulator to optimize an agent's exploration strategies without expensive online reruns.

  • Project
    yc-duan/fastctx

    FastCtx is a framework that lets LLM agents like Codex and Claude Code run across multiple personal machines as a unified “World,” sharing file‑system access and a common action journal while respecting each provider’s usage terms.

  • Project
    openonion/connectonion

    A CLI harness that gives AI agents a unified command-line interface to access and More tools like Gmail, Slack, and browsers, keeping context discovery on-demand and local.