HolmesGPT
Open-source AI agent that investigates Kubernetes incidents and finds root causes.
About
HolmesGPT is an open-source AI SRE agent that aggregates signals from logs, metrics, alerts, and distributed traces, then uses large language models to synthesize evidence and explain infrastructure failures in natural language. It runs a ReAct loop over kubectl, logs, Prometheus, Grafana, and Loki with more than 30 toolsets, operating read-only by design and respecting Kubernetes RBAC.
Maintained by Robusta with contributions from Microsoft, HolmesGPT was accepted into the CNCF Sandbox in October 2025 and is licensed Apache 2.0. It supports BYO LLM including local inference via Ollama for air-gapped environments, integrates with Alertmanager, PagerDuty, OpsGenie, Jira, and Slack, and can open suggested-fix pull requests in Operator mode.
What it can do
- AI root cause analysis for Kubernetes incidents
- 30+ toolsets across kubectl, Prometheus, Grafana, Loki
- Read-only by design with RBAC respect
- Suggested-fix pull requests in Operator mode
- BYO LLM including air-gapped local inference
Setup
Install via Homebrew tap, Helm chart, or Docker; point it at an LLM provider and your cluster.
Prerequisites
Kubernetes cluster access and an LLM API key or local Ollama instance.
Use cases
Works with
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 134; Trending score 82/100 from researcher assessment.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.