AugAgent 🤖

AugAgent is an enterprise-grade multi-agent orchestration framework designed for resilience, observability, and compliance.

✨ Features

• Resilience: Built-in multi-model fallback routing and retries across OpenAI, Ollama, and LiteLLM compatible APIs.
• Persistence: Pluggable state checkpointers (SQLite & ChromaDB) for durable agent memory.
• Extensibility: Native MCP (Model Context Protocol) client wrapper to inject dynamic tools.
• Assurance: OpenTelemetry logging and strict Pydantic token budgets.
• Security: Immutable JSONL audit logging with automatic PII redaction and FastAPI RBAC.

🚀 Quickstart

Installation

pip install augagent[api,telemetry,mcp]

Basic Usage

import asyncio
from augagent.agent import AugAgent
from augagent.models import AgentConfig, LLMConfig

async def main():
    agent = AugAgent(
        config=AgentConfig(
            name="ResearchBot",
            role="Analyst",
            goal="Analyze technical data",
            llm_config=LLMConfig(model="gpt-4o")
        )
    )
    
    result = await agent.run("What are the benefits of MCP?")
    print(result.output)

if __name__ == "__main__":
    asyncio.run(main())

🧪 Testing

AugAgent includes a comprehensive test suite (15+ tests) built with pytest and httpx.MockTransport covering the ReAct loop, token budgets, HITL, streaming, and audit features.

pip install pytest pytest-asyncio httpx
pytest tests/ -v

📖 Documentation

Please refer to our complete API reference and architectural guides.

🤝 Contributing

See CONTRIBUTING.md for how to help out.

🔐 Security

See SECURITY.md for vulnerability reporting.

📄 License

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