AnujaMehrotra/financial-agent — reverse-engineered prompt

Reverse engineered prompt

GitHub

Build me a Python tool that can read a company financial filing, like a 10 K text file, and answer questions about it using AI.

I want it to work like a small multi agent financial research assistant. It should find the most relevant parts of the document, check that the chunks actually match the user’s question, summarize the useful information, and then give a deeper analyst style answer. Use OpenAI for the language model and a local vector search approach for retrieval.

Please set it up so I can put a filing text file in a data folder, add my OpenAI API key in an env file, and run one main Python command to ask a question like “What are Tesla’s key business risks in 2023?” Keep the code modular, with separate agents for retrieval, validation, summarizing, analysis, and user input, and wire the flow together with LangGraph and LangChain.

Want more depth? Deep Reverse