# holmesgpt

> holmesgpt is an open-source agent for cloud and infrastructure. Open-source SRE agent that investigates production incidents and finds root causes. It is available on CLI, self-hosted and Slack.

- **Kind:** Agent
- **Pricing:** Open source
- **Platforms:** CLI, self-hosted, Slack
- **Website:** https://holmesgpt.dev
- **Repository:** https://github.com/holmesgpt/holmesgpt
- **GitHub stars:** 3,916
- **License:** Apache-2.0
- **Categories:** cloud and infrastructure, developer tools
- **Listing updated:** 2026-11-14

## About

Part of the CNCF sandbox, HolmesGPT was created by Robusta.Dev with major contributions from Microsoft. It queries your observability and infrastructure data, such as Prometheus, Grafana, Datadog and Kubernetes, and also any REST API you define, then reasons about what went wrong. It does not require Kubernetes and can work with VMs, bare metal and cloud services.

A newer operator mode runs in the background in Kubernetes, runs scheduled health checks and deployment verification, and posts findings to Slack. With the GitHub integration it can open pull requests with fixes. Alert sources include AlertManager, PagerDuty, OpsGenie and Jira, and findings can be written back. Per-tool memory limits and output budgeting keep large queries from overwhelming the model.

## Features

- **Incident investigation:** Looks across data sources to find the root cause of a problem.
- **Operator mode:** Runs 24/7 in Kubernetes with scheduled health checks and Slack messages.
- **Deep integrations:** Prometheus, Grafana, Datadog, Kubernetes and custom REST toolsets.
- **Alert round-trip:** Fetches alerts from AlertManager, PagerDuty, OpsGenie or Jira and writes findings back.
- **Any LLM provider:** OpenAI, Anthropic, Azure, Bedrock, Gemini and others.
- **Memory-safe execution:** Per-tool memory limits and output budgeting for large datasets.

## Good for

- Diagnose a failing deployment
- Triage a PagerDuty alert with automated evidence gathering
- Verify a new release is healthy
- Open a fix PR for a detected problem

## FAQ

### Does it need Kubernetes?

No. It works with VMs, bare metal and cloud services; only the operator mode runs in Kubernetes.

### Which models can it use?

The README lists OpenAI, Anthropic, Azure, Bedrock, Gemini and more.

### Who maintains it?

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

### Is holmesgpt free?

Yes. holmesgpt is open source, so you can use and self-host it for free.

### What are the best alternatives to holmesgpt?

The closest alternatives to holmesgpt are deepseek-harness, gemini-cli, learn-claude-code, Codewhale and oh-my-pi.

## Alternatives

- [deepseek-harness](https://ai-powered.com/agents/deepseek-harness): deepseek-harness is an open-source agent for multi-agent frameworks.
- [gemini-cli](https://ai-powered.com/agents/gemini-cli): gemini-cli is an open-source agent for ai models and llm tooling.
- [learn-claude-code](https://ai-powered.com/agents/learn-claude-code): learn-claude-code is an open-source hosted agent for ai models and llm tooling.
- [Codewhale](https://ai-powered.com/agents/codewhale): Codewhale is an open-source agent for coding.
- [oh-my-pi](https://ai-powered.com/agents/oh-my-pi): oh-my-pi is an open-source agent for coding.
- [openclaude](https://ai-powered.com/agents/openclaude): openclaude is an open-source agent for coding.
- [qwen-code](https://ai-powered.com/agents/qwen-code): qwen-code is an open-source agent for coding.
- [learn-harness-engineering](https://ai-powered.com/agents/learn-harness-engineering): learn-harness-engineering is an open-source agent for coding.

---
Source: https://ai-powered.com/agents/holmesgpt
