Founder Scraper

AI-powered founder enrichment and scoring tool for VC deal flow. Takes a founder's name, scrapes public data from multiple sources, and outputs a scored founder card analyzed against a configurable investment thesis.

How it works

Founder name → Scrape public data → Enrich profile → Score with Claude → Founder card

Data sources:

• GitHub (repos, languages, stars, activity)
• Crunchbase (funding history, company info, prior companies)
• Twitter/X (bio, followers, recent topics)
• DuckDuckGo (supplementary web context)

Scoring categories (configurable weights):

Category	Default Weight
Founder Quality	30%
Market	25%
Traction	25%
Network	10%
Intangibles	10%

Each category is scored 0-100 with a rationale. The overall score is a weighted average.

Setup

git clone https://github.com/joyjcai88-collab/founder-scraper.git
cd founder-scraper
pip install -e .
cp .env.example .env

Add your Anthropic API key to .env:

ANTHROPIC_API_KEY=sk-ant-...

Usage

# Basic usage
python main.py "Tobi Lutke"

# With company name for better accuracy
python main.py "Tobi Lutke" --company "Shopify"

# Short flag
python main.py "Patrick Collison" -c "Stripe"

Output

The CLI outputs a formatted founder card with:

• Overall score (color-coded green/yellow/red)
• Per-category breakdown with bar charts and rationale
• Thesis fit summary
• Key risks and flags
• Source links

Project Structure

founder-scraper/
├── main.py              # CLI entry point
├── scraper/
│   ├── github.py        # GitHub REST API
│   ├── crunchbase.py    # Crunchbase page scraper
│   ├── twitter.py       # Twitter/X via Nitter
│   ├── web_search.py    # DuckDuckGo search
│   └── safety.py        # Input sanitization, SSRF protection
├── analyzer/
│   ├── enricher.py      # Concurrent scraper orchestration
│   └── scorer.py        # Claude API scoring engine
├── models/
│   └── founder.py       # Pydantic data models
└── config/
    └── thesis.py        # Investment thesis templates

Security

• API keys excluded from version control via .gitignore
• Input sanitization on all user-provided strings
• SSRF protection blocks requests to private/internal IPs
• Scraped content is delimited and treated as untrusted data in LLM prompts
• All dependencies are version-pinned

Requirements

• Python 3.9+
• Anthropic API key