TraceRoot is an open-source debugging platform that helps engineers fix production issues 10x faster by combining structured traces, logs, and source code context with AI-powered analysis. Contributing | TraceRoot.AI | Community | SDK | Documentation Join us (Discord) in pushing the boundaries of debugging with AI agents. Please 🌟 Star TraceRoot on GitHub and be instantly notified of new releases. TraceRoot Framework Design Principles 🤖 Intelligence The framework enables multi-agent systems to continuously evolve by interacting with environments. The framework enables real-time tracing and logging to your applications. 🧠 Structured Information The framework enables utilizing structured loggings and tracing data to improve the performance of AI agents. 💻 Integration The framework enables integrating with other sources and tools, such as GitHub, Notion, etc. This provides a seamless experience for you to use the framework in your applications. 😊 Developer Friendly We also provide a Cursor like interface but specialized for debugging and tracing. You can select the logs and traces you are interested in and ask the framework to help you with the analysis. Why Use TraceRoot for Your Applications? We are a community-driven collective comprising over multiple engineers and researchers dedicated to advancing frontier engineering and research in using Multi-Agent Systems to help not only human but also AI agents on debugging, tracing, and root cause analysis. ✅ Multi-Agent System Multi-Agent system that can be used to solve complex tasks. ✅ Real-Time Tracing and Logging Enable real-time tracing and logging to your applications. ✅ Structured Logging Enable structured logging to your applications, which allows better performance of AI agents. ✅ Integration with Multiple Resources Integrate with other sources and tools, such as GitHub, Notion, etc. ✅ Developer Friendly We provide a Cursor like interface but specialized for debugging and tracing. AI Agent Framework Here is an overview for our AI Agent Framework: Context Model Chunking Explainability Please checkout the README.md in the rest/agent directory for more details. Getting started with TraceRoot TraceRoot Cloud (Recommended) The fastest and most reliable way to get started with TraceRoot is signing up for free to TraceRoot Cloud for a 7 day trial. You will have 150k trace + logs storage with 30d retentions, 1.5M LLM tokens, and AI agent with chat mode. Self-hosting the open-source deploy (Advanced) Installation You can install the latest version of TraceRoot with the following command: Install the dependencies locally: git clone https://github.com/traceroot-ai/traceroot.git cd traceroot # Create and activate a virtual environment python3.11 -m venv venv source venv/bin/activate # Install TraceRoot with dependencies excluding optional dependencies pip install --upgrade pip pip install -e . Local Usage For local usage, all of your data will be stored locally. Run the below command to intialize environment variables. source .env.development You can use the TraceRoot framework locally by following the README.md in the ui directory and README.md in the rest directory. Also, you can build the latest docker image and run the docker container by following the README.md in the docker directory. This will start the UI at http://localhost:3000 and the API at http://localhost:8000. Before using the TraceRoot framework, you need to setup the Jaeger docker container at first. It will be used to store the traces and logs and capture the traces and logs from our SDK which is integrated with your applications. docker run -d --name jaeger \ -e COLLECTOR_OTLP_ENABLED=true \ -p 16686:16686 \ -p 14268:14268 \ -p 14250:14250 \ -p 4317:4317 \ -p 4318:4318 \ cr.jaegertracing.io/jaegertracing/jaeger:2.8.0 In local mode, the first step is to go to the integration page and connect with your GitHub account (optional) with your GitHub token. You also need to put your OpenAI API key in the integration page. Setting up TraceRoot Our project is built on top of the TraceRoot SDK. You need to use our SDK to integrate with your applications by pip install traceroot==0.0.4a7 To use the local mode of the TraceRoot SDK, you need create a .traceroot-config.yaml file in the root directory of your project with following content: local_mode : true service_name : " your-service-name " github_owner : " your-github-owner " github_repo_name : " your-github-repo-name " github_commit_hash : " your-github-commit-hash " As mentioned above, you need to setup the Jaeger docker container at first before let the TraceRoot SDK capture the traces and logs from your applications. For more details or the SDK usage and examples, please checkout this Quickstart. Citation If you find our exploratory TraceRoot useful in your research, please consider citing: