v1.0.0
AgentNeo v1.0.0
We're excited to announce the initial release of AgentNeo, an open-source Agentic AI Application Observability, Monitoring, and Evaluation Framework.
What's New
- Initial release of AgentNeo core functionality
- Tracing capabilities for LLM calls, agents, and tools
- Interactive dashboard for visualization and analysis
- SQLite and JSON-based data storage
- Project management features
- Execution graph visualization
Features
- Trace LLM Calls: Monitor calls from providers like OpenAI and LiteLLM
- Trace Agents and Tools: Instrument and monitor your custom agents and tools
- Monitor Interactions: Track tool and agent interactions
- Collect Metrics: Gather data on token usage, costs, and execution time
- Flexible Data Storage: Use SQLite databases and JSON log files
- Simple Instrumentation: Utilize decorators for easy code instrumentation
- Interactive Dashboard: Visualize trace data and execution graphs
- Project Management: Handle multiple projects within the framework
- Evaluation Tools: Assess AI agent performance
Installation
pip install agentneo
Quick Start
from agentneo import AgentNeo, Tracer, launch_dashboard
neo_session = AgentNeo(session_name="my_session")
neo_session.create_project(project_name="my_project")
tracer = Tracer(session=neo_session, log_file_path="trace.json")
tracer.start()
# Instrument your code here
tracer.stop()
launch_dashboard(port=3000)
Requirements
- Python 3.8+
- Node.js 14+
- npm 6+ or yarn 1.22+
Documentation
For detailed usage instructions, visit our official documentation.
What's Next
We're actively working on:
- Support for additional LLMs
- Integration with AutoGen, CrewAI, and Langraph
- Enhanced logging capabilities
- Custom agent orchestration support
Contributing
We welcome contributions! Please check our contribution guidelines to get started.
License
AgentNeo is released under the (https://github.com/raga-ai-hub/agentneo/blob/main/LICENSE).