structured-report 0.1.1

Transform text into structured thought models for deep learning

Deep Thinking Report API - MVP

A streamlined API that transforms text into structured thought models for learning and reflection. Built with FastAPI and supports multiple LLM providers.

Features

• POST /v1/report/insight: Generate structured analysis from text
• Dual Model Support:
  • OpenAI GPT-4o-mini: Paid model with higher reasoning capabilities
  • DeepSeek: Free model with good performance
• Depth Levels: Support for light (1), standard (2), and deep (3) analysis
• Structured Output: Consistent JSON response format
• Docker Support: Easy deployment with Docker
• Simplified Architecture: Clean 134-line codebase for easy maintenance

Installation

Option 1: Install from PyPI

pip install structured-report

Run the server:

structured-report

Or run programmatically:

from structured_report.main import app
import uvicorn
uvicorn.run(app, host="0.0.0.0", port=8000)

Option 2: Run from source

Quick Start

1. Clone and setup:

   git clone <repository-url>
   cd Report_API
   

2. Configure environment:

   # Create .env file with your API keys
   echo "OPENAI_API_KEY=your_openai_api_key_here" > .env
   echo "DEEPSEEK_API_KEY=your_deepseek_api_key_here" >> .env
   

   Get API keys from:

   • OpenAI: https://platform.openai.com/api-keys
   • DeepSeek: https://platform.deepseek.com/

3. Run with Docker:

   docker-compose up --build
   

4. Test the API:

   curl -X POST "http://localhost:8000/v1/report/insight" \
     -H "Content-Type: application/json" \
     -d '{
       "content": "The future of AI will be dominated by general intelligence systems that can learn and adapt like humans.",
       "model": "deepseek",
       "depth_level": 2
     }'
   

API Usage

Authentication

This API requires an API Key for all requests. Get your API Key first:

Get API Key:

curl -X POST "http://localhost:8000/api-keys" \
  -H "Content-Type: application/json" \
  -d '{"name": "your-name", "email": "[e-mail]"}'

Use API Key in requests:

curl -X POST "http://localhost:8000/v1/report/insight" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{"content": "your text", "model": "deepseek", "depth_level": 2}'

Rate Limiting

This API implements rate limiting to ensure fair usage:

• API Key Creation: 5 requests per minute per IP address
• Insight Generation: 10 requests per minute per IP address
• Debug Endpoints: 10 requests per minute per IP address

When rate limited, you'll receive a 429 Too Many Requests response. Rate limits reset every minute.

Check current rate limits:

curl http://localhost:8000/rate-limits

Endpoint: POST /v1/report/insight

Request:

{
  "content": "Your text content here",
  "model": "openai",  // or "deepseek"
  "depth_level": 2    // 1=light, 2=standard, 3=deep
}

Model Options:

• "openai": GPT-4o-mini (paid, higher reasoning quality)
• "deepseek": DeepSeek Chat (free, good performance)

Response:

{
  "core_claim": "Extracted main thesis",
  "supporting_arguments": [
    {
      "claim": "Supporting point",
      "evidence": "Evidence from text",
      "logic_type": "Type of reasoning"
    }
  ],
  "assumptions": ["Implicit assumptions"],
  "speaker_position": {
    "role": "Speaker's role",
    "potential_bias": "Potential biases"
  },
  "my_evaluation": "[user will fill]",
  "personal_impact": "[user will fill]",
  "meta_principle": "Transferable principle",
  "falsifiability": ["Conditions where claim fails"],
  "keywords": [
    {
      "term": "Key term",
      "definition": "Definition"
    }
  ],
  "related_insights": ["Related concepts"],
  "meta_questions": ["Questions for further thinking"]
}

Development

Run locally:

pip install -r requirements.txt
uvicorn app:app --reload

API Documentation: Visit http://localhost:8000/docs for interactive API documentation.

Architecture

Clean single-file design with 134 lines of code:

• app.py: Complete FastAPI application with Pydantic models
• Supports both OpenAI and DeepSeek models
• Simplified prompt engineering and JSON parsing
• Docker-ready deployment

MVP Limitations

• Basic error handling
• Simple API Key authentication (no user accounts)
• Rate limiting per IP address (5-10 requests/minute)
• No caching
• No advanced features (batch processing, web UI)

Next Steps (Future Versions)

• User authentication & API keys
• Advanced rate limiting (per API key quotas)
• Response caching
• Web UI interface
• Batch processing
• Advanced prompt engineering
• Knowledge graph visualization