AI-Kline is a Python-based A-share stock analysis tool that merges traditional technical indicators with AI-driven forecasting. It targets individual investors and researchers, offering comprehensive analysis via CLI, Web UI, and an MCP server, leveraging historical data, financial news, and advanced AI predictions.

How It Works

The tool fetches A-share historical trading data, financial information, and news using AKShare. It computes standard technical indicators (MA, MACD, KDJ, RSI, Bollinger Bands) and visualizes them with K-line charts. A key feature is its AI analysis module, employing multimodal AI models (e.g., Qwen-VL-MAX via Dashscope, or potentially Gemini) to analyze integrated data and forecast future stock trends.

Quick Start & Requirements

• Installation: Clone the repo; pip install -r requirements.txt.
• Prerequisites: Python 3.8+, API keys for AI models (e.g., OpenAI, Dashscope, Gemini). Requires a .env file for API_KEY, BASE_URL, MODEL_NAME.
• Usage:
  • CLI: python main.py --stock_code <CODE> --period <PERIOD>
  • Web UI: python web_app.py (access at http://localhost:5000)
  • MCP Server: uv run mcp_server.py (connect via http://localhost:8000/mcp)
• Links: No specific quick-start or demo links provided.

Highlighted Details

• Data acquisition via AKShare for A-shares (trading, financial, news).
• Supports numerous technical indicators (MA, MACD, KDJ, RSI, Bollinger Bands).
• Offers static (PNG) and interactive (HTML) chart visualizations.
• Integrates multimodal AI for predictive analysis, with Web UI and MCP Server interfaces.

Maintenance & Community

Part of the "QuantML" open-source initiative. Specific maintenance, contributor, or community channel details are not provided. Commercial use requires direct contact with QuantML.

Licensing & Compatibility

An open-source project from QuantML requiring attribution. Commercial use requires contact. The specific license type is not stated, necessitating clarification for commercial compatibility.

Limitations & Caveats

Intended for learning/research only, not investment advice. AI predictions are not guaranteed. A contradiction exists regarding required API keys: .env example uses Dashscope (qwen-vl-max), while a caveat mentions Gemini. Data acquisition depends on AKShare, subject to network/source limitations.