How do I run a full end-to-end stock analysis locally?

[ryanfu123]

I want to run a comprehensive analysis on a single ticker (e.g., NVDA) and get the full output — DCF model, news intelligence, and analyst report. What's the fastest path from clone to results?

[zanwenfu]

Prerequisites: Python 3.11+, and API keys for OpenAI or Anthropic + SerpAPI (for news scraping). MongoDB is optional.

# Clone and set up
git clone https://github.com/Agentic-Analyst/stock-analyst.git
cd stock-analyst
conda create -n stock-analyst python=3.11 -y && conda activate stock-analyst
pip install -r requirements.txt
pip install --upgrade --force-reinstall git+https://github.com/Agentic-Analyst/vynn-core.git

# Configure API keys
cp .env.example .env
# Edit .env → set OPENAI_API_KEY or ANTHROPIC_API_KEY + SERPAPI_API_KEY

# Run a full comprehensive analysis
python main.py \
  --pipeline chat \
  --email your@email.com \
  --timestamp $(date +%Y%m%d_%H%M%S) \
  --llm claude-3.5-sonnet \
  --user-prompt "Analyze NVDA comprehensively with focus on AI chip market"

This triggers all 4 agents (financial data → DCF model → news intelligence → report). Takes ~6-7 minutes. Outputs land in data/{email}/NVDA/{timestamp}/ — you'll get a 10-tab Excel DCF, screening JSON with catalysts/risks, and a professional analyst report in markdown.

If you just want the DCF model without news, use --pipeline financial-model with --ticker NVDA instead, which finishes in under 15 seconds.