shivam1423/Financial-RAG-System — reverse-engineered prompt

Reverse engineered prompt

GitHub

Build me a local AI financial analyst that I can use from Cursor chat. I want to ask normal questions like “show Tesla’s performance for the last 3 months” or “compare Apple and Microsoft over the past year” and have it turn that into real stock market analysis with Python charts.

Use a local Deepseek R1 model through Ollama, coordinate the work with CrewAI agents, and expose it as an MCP server so Cursor can call it directly. It should fetch stock data, analyze prices and trading volume, compare tickers when I ask, and generate clear matplotlib visualizations.

Please make the setup easy for someone on Python 3.12, with uv support, a server file Cursor can run as an MCP tool, and a simple notebook or example showing the workflow. Include clear instructions for installing Ollama, pulling the model, syncing dependencies, and adding the MCP server in Cursor. Look up current docs online if needed.

Want more depth? Deep Reverse