# OpenSRE

> OpenSRE is an open-source (Apache 2.0) AI SRE agent framework that automates incident investigation and root cause analysis.

OpenSRE connects to your existing observability and infrastructure stack (Datadog, Grafana, Prometheus, CloudWatch, Kubernetes, and 60+ integrations in total), investigates alerts autonomously — gathering context, testing hypotheses in parallel, and reasoning over evidence — and delivers root-cause reports with recommended fixes to Slack, Discord, or your incident tooling. It runs locally, on-prem, or in the cloud, with your choice of LLM provider.

## Docs

- [Quickstart](https://opensre.com/docs/quickstart.md): Install OpenSRE and run your first investigation in minutes
- [Installation](https://opensre.com/docs/install.md): Local, on-prem, and enterprise installation paths
- [How an investigation works](https://opensre.com/docs/how-investigations-work.md): What OpenSRE does between alert intake and the root-cause report
- [Architecture](https://opensre.com/docs/ARCHITECTURE.md): Investigation pipeline and system architecture
- [Integrations overview](https://opensre.com/docs/integrations-overview.md): Observability, cloud, database, and messaging integrations
- [LLM providers](https://opensre.com/docs/llm-providers.md): Supported LLM APIs and CLIs, and how to switch between them
- [FAQ](https://opensre.com/docs/faq.md): Frequently asked questions about OpenSRE

## Comparisons

- [How OpenSRE fits your stack](https://opensre.com/docs/comparisons/overview.md): Where OpenSRE sits relative to observability tools, orchestrators, and cloud-native monitoring

## Benchmarks

- [CloudOpsBench results](https://opensre.com/docs/cloudopsbench.md): OpenSRE+LLM vs LLM-alone baselines on the 452-scenario CloudOpsBench corpus
- [Benchmarks](https://opensre.com/benchmarks): Incident-investigation benchmark reports

## Project

- [GitHub repository](https://github.com/Tracer-Cloud/opensre): Apache 2.0 source code
- [Blog](https://opensre.com/resources): Technical articles on incident investigation and root cause analysis

## Optional

- [Full docs index](https://opensre.com/docs/llms.txt): Machine-readable index of every documentation page
- [Full docs content](https://opensre.com/docs/llms-full.txt): Complete flattened documentation for LLM consumption