Financial Analysis Agent

Streamlit + CrewAI application that runs a 5-agent hierarchical crew against an NSE India stock symbol and returns an investment write-up.

How to Run

1. Install dependencies:

   pip install -r requirements.txt

2. Set environment variables: Create a .env file in the root of the project:

   GEMINI_API_KEY=YOUR_GEMINI_API_KEY
   SERPER_API_KEY=YOUR_SERPER_API_KEY
   LANGSEARCH_API_KEY=YOUR_LANGSEARCH_API_KEY
   PASSKEY=YOUR_PASSKEY
   

   Optional tuning variables are documented in src/config.py.

3. Run the application:

   streamlit run app.py

Architecture

app.py                     # Streamlit UI
stock_analysis.py          # analyse_stock (UI) + run_analysis (worker)
src/
  config.py                # All configuration
  stocks.py                # Stock symbol validation
  tools/                   # CalculatorTool, LangsearchTool
  crew/                    # CrewAI agents, tasks, and runner
  storage/                 # SQLite (WAL mode) — cache, queue, in-flight tracking
  workers/                 # Background queue processing thread
data/
  stock.csv                # NSE India symbols

Requests are enqueued and processed asynchronously by a background worker thread — the UI never blocks on crew execution. Results are cached for 3 days (configurable).

Documentation

See docs/ for detailed architecture, security, concurrency, and testing documentation.