teraace-tracker 1.1.8

A comprehensive Python library for tracking AI agent events across LangChain, CrewAI, and AutoGPT frameworks

Teraace Agentic Tracker

A comprehensive Python library for tracking AI agent events across 20 major agentic frameworks and sending them to the Teraace API for monitoring and analysis.

🚀 Features

• 20 Framework Support: LangChain, CrewAI, AutoGPT, Swarm, LlamaIndex, AutoGen, Phidata, BabyAGI, MetaGPT, TaskWeaver, CAMEL, AgentGPT, SuperAGI, Semantic Kernel, Haystack, Rasa, PydanticAI, DSPy, Mirascope, Instructor
• Unified Tracking: Consistent event tracking across all frameworks
• Production Ready: Buffered delivery, async API client, graceful shutdown
• Thread-Safe: Safe for multi-threaded applications
• Extensible: Easy to add new framework integrations

📦 Teraace Account

Create your Teraace account at https://app.teraace.com/signup

📦 Installation

pip install teraace-tracker

⚡ Quick Start

1. Configuration

Create a .env file:

TERAACE_API_KEY=your_api_key_here
TERAACE_BUFFER_SIZE=20
TERAACE_API_ENDPOINT=https://api.teraace.com/agents
TERAACE_REQUEST_TIMEOUT=30
TERAACE_MAX_RETRIES=3

2. Basic Usage

from teraace_tracker import LangChainTracker

# Initialize tracker
tracker = LangChainTracker()

# Use with your LangChain agent
agent = initialize_agent(
    tools=your_tools,
    llm=llm,
    callbacks=[tracker]  # Add tracker as callback
)

# Use agent normally - events are tracked automatically
result = agent.run("What is 2 + 2?")

3. Other Frameworks

• AGENTS.md - Comprehensive examples for all 20 frameworks

# CrewAI
from teraace_tracker import CrewAITracker
tracker = CrewAITracker()
tracker.track_crew_execution(crew)

# OpenAI Swarm
from teraace_tracker import SwarmTracker
tracker = SwarmTracker()

# LlamaIndex
from teraace_tracker import LlamaIndexTracker
tracker = LlamaIndexTracker()

# AutoGen
from teraace_tracker import AutoGenTracker
tracker = AutoGenTracker()

# And 11 more frameworks...

🎯 Supported Frameworks

Framework	Status	Use Case
LangChain	✅	General-purpose agent framework
CrewAI	✅	Multi-agent collaboration
AutoGPT	✅	Autonomous task execution
OpenAI Swarm	✅	Multi-agent orchestration
LlamaIndex	✅	Data-centric agents & RAG
AutoGen	✅	Multi-agent conversations
Phidata	✅	Production AI assistants
BabyAGI	✅	Autonomous task prioritization
MetaGPT	✅	Software development teams
TaskWeaver	✅	Code-first stateful agents
CAMEL	✅	Communicative agent societies
AgentGPT	✅	Web-based autonomous agents
SuperAGI	✅	Open-source agent infrastructure
Semantic Kernel	✅	Microsoft AI orchestration
Haystack	✅	NLP/RAG with agents
Rasa	✅	Conversational AI
PydanticAI	✅	Production-grade AI applications
DSPy	✅	Prompt optimization framework
Mirascope	✅	Type-safe LLM toolkit
Instructor	✅	Structured data extraction

📊 What Gets Tracked

• Agent Lifecycle: Creation, execution, completion
• Tool Usage: Function calls and results
• Memory Operations: Read, write, update operations
• Multi-Agent Interactions: Handoffs and collaboration
• Code Generation: Generated code and execution
• Planning & Reasoning: Agent thought processes
• Resource Management: File and data operations

🔧 Configuration

Environment Variable	Default	Description
TERAACE_API_KEY	Required	Your Teraace API key
TERAACE_BUFFER_SIZE	20	Events to buffer before sending
TERAACE_API_ENDPOINT	https://api.teraace.com/agent-events	API endpoint

📚 Documentation

• AGENTS.md - Comprehensive examples for all 20 frameworks
• API Reference - Detailed API documentation
• Configuration Guide - Advanced configuration options

🛠️ Development

git clone https://github.com/teraace/teraace-tracker
cd teraace-tracker
pip install -e .[dev]
pytest

🤝 Contributing

We welcome contributions! Please see our Contributing Guide for details.

📄 License

MIT License - see LICENSE file for details.

🆘 Support

• Documentation: docs.teraace.com
• Email: [e-mail]
• Issues: GitHub Issues