OpenContext

AI-powered company context analysis using Google Gemini

OpenContext extracts comprehensive company information from any website URL using Google's Gemini AI. Perfect for lead research, competitive analysis, content creation, and business intelligence.

Features

• AI-Powered Analysis using Google Gemini with Google Search grounding
• Visual Identity Extraction for consistent image generation (brand colors, style, mood)
• Blog Authors extraction from existing articles
• Writing Persona (voice_persona) tailored to ICP for content creation
• CLI & REST API - Use from terminal or integrate via HTTP
• Structured JSON output matching openblog schema
• Modular Architecture with shared GeminiClient

What's New in v3.0

• REST API: FastAPI server with async job support
• Visual Identity: Brand colors, design elements, typography, image style prompts
• Blog Image Examples: Analyzes existing blog images for style reference
• Authors Extraction: Finds real blog authors with profiles
• Shared GeminiClient: Unified client with retry logic, URL Context, Google Search
• Prompt Files: Prompts stored in text files for easy iteration
• Full Parity with openblog/stage 1 schema

Installation

git clone https://github.com/federicodeponte/opencontext.git
cd opencontext
pip install -e .

Configuration

Set your Gemini API key:

# Option 1: Environment variable
export GEMINI_API_KEY=your_key_here

# Option 2: Use configure command
opencontext config

Get a Gemini API key at aistudio.google.com

Usage

# Analyze a company website
opencontext analyze https://example.com

# Verbose output (shows all details)
opencontext analyze https://example.com -v

# Save output to specific file
opencontext analyze https://example.com -o company.json

# Check configuration
opencontext check

# Show help
opencontext --help

REST API

Start the API server:

# Install API dependencies
pip install -e ".[api]"

# Start server
uvicorn api:app --reload --port 8000

# API docs at http://localhost:8000/docs

API Endpoints

Method	Endpoint	Description
GET	/	Health check
POST	/api/v1/analyze	Analyze company (sync, blocking)
POST	/api/v1/jobs	Start analysis job (async)
GET	/api/v1/jobs	List all jobs
GET	/api/v1/jobs/{id}	Get job status/result
DELETE	/api/v1/jobs/{id}	Delete job

Example API Usage

# Synchronous analysis (blocking)
curl -X POST http://localhost:8000/api/v1/analyze \
  -H "Content-Type: application/json" \
  -d '{"url": "https://example.com"}'

# Async job (non-blocking)
curl -X POST http://localhost:8000/api/v1/jobs \
  -H "Content-Type: application/json" \
  -d '{"url": "https://example.com"}'

# Check job status
curl http://localhost:8000/api/v1/jobs/{job_id}

Project Structure

opencontext/
├── cli.py                      # CLI entry point
├── api.py                      # FastAPI REST API
├── opencontext/                # Main package
│   ├── __init__.py
│   ├── models.py               # Pydantic models (CompanyContext, VoicePersona, etc.)
│   ├── opencontext.py          # Core analysis logic
│   └── prompts/
│       └── opencontext.txt     # Analysis prompt (170 lines)
├── shared/                     # Shared components
│   ├── __init__.py
│   ├── constants.py            # GEMINI_MODEL config
│   ├── gemini_client.py        # Unified Gemini client
│   └── prompt_loader.py        # Load prompts from files
├── pyproject.toml
├── requirements.txt
├── CLAUDE.md
└── README.md

What Gets Extracted

Company Basics

• Company name, website, industry
• Products/services offered
• Target audience (ICP)
• Main competitors
• Brand tone and voice

Content Strategy

• Customer pain points
• Value propositions
• Use cases
• Content themes

Writing Persona (voice_persona)

• ICP profile description
• Voice style for content
• Language style (formality, complexity, perspective)
• Sentence patterns
• Do's and don'ts
• Example phrases
• Opening styles

Visual Identity (for Image Generation)

• Brand colors (hex codes)
• Secondary colors
• Visual style (minimalist, bold, corporate, etc.)
• Design elements (gradients, icons, illustrations)
• Typography style
• Mood
• Image style prompt (for AI image generation)
• Blog image examples with descriptions
• Elements to avoid

Blog Authors

• Full name
• Job title/role
• Bio
• Profile image URL
• LinkedIn/Twitter URLs

Example Output

{
  "company_name": "Example Corp",
  "company_url": "https://example.com",
  "industry": "Technology",
  "description": "A leading provider of...",
  "products": ["Product A", "Product B"],
  "target_audience": "Tech-savvy SMBs looking for...",
  "competitors": ["Competitor 1", "Competitor 2"],
  "tone": "professional",
  "voice_persona": {
    "icp_profile": "CTOs and engineering leaders at mid-sized companies",
    "voice_style": "Direct and technical, focusing on practical implementation",
    "language_style": {
      "formality": "professional",
      "complexity": "technical",
      "sentence_length": "mixed",
      "perspective": "peer-to-peer"
    },
    "do_list": ["Use code examples", "Show benchmark data"],
    "dont_list": ["Avoid marketing fluff", "Skip obvious statements"]
  },
  "visual_identity": {
    "brand_colors": ["#1E40AF", "#3B82F6"],
    "visual_style": "Modern minimalist with tech focus",
    "mood": "Professional, innovative, trustworthy",
    "image_style_prompt": "Modern minimalist illustration with blue gradient...",
    "blog_image_examples": [
      {
        "url": "https://example.com/blog/hero.jpg",
        "description": "Abstract blue gradient with geometric shapes",
        "image_type": "hero"
      }
    ]
  },
  "authors": [
    {
      "name": "John Smith",
      "title": "CTO",
      "bio": "Leading engineering at Example Corp",
      "linkedin_url": "https://linkedin.com/in/johnsmith"
    }
  ]
}

Using as a Library

import asyncio
from opencontext import get_company_context, CompanyContext

async def main():
    # Get context with fallback
    context, ai_called = await get_company_context("https://example.com")

    print(f"Company: {context.company_name}")
    print(f"Industry: {context.industry}")
    print(f"Colors: {context.visual_identity.brand_colors}")
    print(f"Authors: {[a.name for a in context.authors]}")

    # Export to dict
    data = context.model_dump()

asyncio.run(main())

Integration with OpenBlog

This schema is fully compatible with openblog Stage 1:

# In openblog, use OpenContext output directly
from opencontext import get_company_context

context, _ = await get_company_context(company_url)
# Pass to Stage 2 for blog generation with consistent visual identity

Environment Variables

Variable	Description
GEMINI_API_KEY	Google Gemini API key
GEMINI_MODEL	Model to use (default: gemini-2.0-flash)

License

MIT License - see LICENSE for details.