Bluelabel Autopilot 🚀

MCP-native agent orchestration platform for intelligent document processing with real-time capabilities.

Overview

Bluelabel Autopilot is a high-performance content processing system that:

• Processes emails with PDF attachments in real-time
• Generates executive summaries using AI agents
• Provides REST API with WebSocket updates
• Features a unified workflow engine architecture
• Achieves <3 second processing for typical documents
• Follows Model Context Protocol (MCP) standards

🎯 Key Features

• Unified Architecture: Single adapter for multiple workflow engines
• Real-time Updates: WebSocket streaming of processing progress
• High Performance: <100ms API response, <3s PDF processing
• Zero Breaking Changes: Backward compatible with existing workflows
• Production Ready: Full test coverage, monitoring, and error handling

Architecture

bluelabel-autopilot/
├── agents/                 # Core agent implementations
│   ├── base_agent.py      # Base class with MCP-compliant I/O
│   ├── agent_models.py    # Shared data models
│   ├── digest_agent.py    # Content digest generation
│   └── ingestion_agent.py # URL and PDF content processing
├── prompts/               # YAML prompt templates
│   └── contentmind/       # Content processing prompts
├── runner/                # CLI and execution
│   └── cli_runner.py      # Command-line interface
└── storage/               # File-based data storage
    └── knowledge/         # Processed content storage

🚀 Quick Start

Installation

# Clone the repository
git clone https://github.com/yourusername/bluelabel-autopilot.git
cd bluelabel-autopilot

# Install dependencies (Python 3.8+)
pip install -r requirements.txt

# Install API dependencies
cd apps/api
pip install -r requirements.txt
cd ../..

Running the System

1. Start the API Server:

cd apps/api
python main.py
# API available at http://localhost:8000
# WebSocket at ws://localhost:8000/ws
# API docs at http://localhost:8000/docs

2. Run the Demo:

python demo/email_to_summary.py

3. Test the Integration:

# Open in browser
open http://localhost:8000/static/test.html

Usage

CLI Commands

Digest Agent

# Generate a digest from stored summaries
python runner/cli_runner.py digest --format markdown

# Add a new summary
python runner/cli_runner.py add-summary "TechCrunch Article" "AI startup raises $50M..." --url "https://example.com"

# Run digest agent with custom JSON input
python runner/cli_runner.py run digest '{"task_id": "test", "content": {"action": "generate_digest"}}'

Ingestion Agent

# Process a URL
python runner/cli_runner.py run ingestion '{
    "task_id": "test-url",
    "task_type": "url",
    "content": {
        "url": "https://example.com/sample-article"
    }
}'

# Process a PDF
python runner/cli_runner.py run ingestion '{
    "task_id": "test-pdf",
    "task_type": "pdf",
    "content": {
        "file_path": "path/to/document.pdf"
    }
}'

# Using sample input files
python runner/cli_runner.py run ingestion --input tests/sample_url_input.json
python runner/cli_runner.py run ingestion --input tests/sample_pdf_input.json

Common Options

# Specify custom storage paths
python runner/cli_runner.py run ingestion --input tests/sample_url_input.json \
    --storage-path ./custom/knowledge \
    --temp-path ./custom/temp

Python API

from agents.digest_agent import DigestAgent
from agents.ingestion_agent import IngestionAgent
from agents.base_agent import AgentInput

# Create agents
digest_agent = DigestAgent()
ingestion_agent = IngestionAgent(
    storage_path=Path("./data/knowledge"),
    temp_path=Path("./data/temp")
)

# Process URL content
url_input = AgentInput(
    task_id="url-001",
    task_type="url",
    source="api",
    content={"url": "https://example.com"}
)
url_result = await ingestion_agent.process(url_input)

# Process PDF content
pdf_input = AgentInput(
    task_id="pdf-001",
    task_type="pdf",
    source="api",
    content={"file_path": "path/to/document.pdf"}
)
pdf_result = await ingestion_agent.process(pdf_input)

# Generate digest
digest_input = AgentInput(
    task_id="digest-001",
    source="api",
    content={"action": "generate_digest", "format": "markdown"}
)
digest_result = await digest_agent.process(digest_input)

Agents

DigestAgent

Generates formatted digests from stored content summaries.

Features:

• Multiple output formats (Markdown, HTML, JSON)
• File-based summary storage
• Chronological ordering
• Source attribution

Example Output:

# Content Digest

Generated on: 2025-01-23

## Summary 1
**Source:** TechCrunch Article  
**Date:** 2025-01-23T10:30:00  
AI startup raises $50M for revolutionary content processing...

Read more

IngestionAgent

Processes content from URLs and PDFs.

Features:

• URL content extraction
• PDF text extraction
• Metadata extraction
• Content validation
• File-based storage

Example Output:

Processing Results:
------------------
Task ID: test-url
Status: success
Duration: 1234ms

Content Details:
Content ID: url_abc123
Content Type: url
Content Length: 12345 characters

Metadata:
title: Sample Article
author: John Doe
date: 2025-01-23

Prompt Templates

The prompts/contentmind/ directory contains YAML templates for:

• summarization.yaml - General content summarization
• event_summary.yaml - Meeting and event summaries
• technical_analysis.yaml - Technical document analysis

Templates use Jinja2 syntax and can be customized for different use cases.

Development

Adding New Agents

1. Create a new file in agents/
2. Inherit from BaseAgent
3. Implement the process() method
4. Add to CLI runner registry

from agents.base_agent import BaseAgent, AgentInput, AgentOutput

class MyAgent(BaseAgent):
    agent_id = "my-agent"
    
    async def process(self, input_data: AgentInput) -> AgentOutput:
        # Your implementation here
        return AgentOutput(
            task_id=input_data.task_id,
            status="completed",
            result="Result"
        )

Running Tests

# Run all tests
pytest

# Run with coverage
pytest --cov=agents --cov=runner

Migration from AIOS-V2

This repository represents a clean extraction of core components from the legacy AIOS-V2 system:

Key Changes:

• Removed PostgreSQL dependencies
• Simplified agent architecture
• Eliminated complex abstractions
• File-based storage instead of database
• MCP-compliant input/output models
• Modernized Python async/await patterns

Preserved Components:

• Core agent logic
• Prompt templates (simplified)
• Content processing workflows

Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• Original AIOS-V2 team for the foundational concepts
• Model Context Protocol (MCP) for standardization guidelines
• The agent-comms-mvp project for architectural patterns