the30top
HO

HolmesGPT

by HolmesGPT (CNCF sandbox project)

Open source SRE agent for investigating production incidents

Reviewed by the the30top research team · Last reviewed 2026-10-27

Overview

HolmesGPT is an open source SRE agent, originally created by Robusta.Dev with major contributions from Microsoft, and now a CNCF sandbox project. It uses an agentic loop to query live observability data across tools such as Datadog, Prometheus, Grafana, Kubernetes and Splunk, and works with providers including OpenAI, Anthropic, Azure, Bedrock and Gemini. By design it has read-only access and respects RBAC. You supply your own model access and run it yourself, so there is no vendor support or SLA.

Our verdict

The leading open source option for SRE investigation, with real community and foundation backing. Budget engineering time for setup.

Best for

Platform teams that want a free, inspectable SRE agent they can run in their own environment.

Avoid if

You need a managed service with support and certifications.

Pros

  • Free and Apache 2.0 licensed
  • Read-only by design
  • Broad data source and model provider support

Cons

  • You run and maintain it yourself
  • No vendor support or compliance reports
  • Model costs are on you

Key capabilities

Agentic loop over live observability data24/7 background operator modeBidirectional alert integrationsRead-only access that respects RBACMany data source integrationsMultiple LLM providers

Frequently asked questions

Is HolmesGPT free?

Yes. It is open source under the Apache 2.0 license. You pay for the model provider you connect.

Can it change my systems?

The project says it has read-only access by design and respects RBAC permissions.

Who maintains it?

It was created by Robusta.Dev with major contributions from Microsoft and is a CNCF sandbox project.

More AI DevOps & Incident Response Agents