Open source agent that investigates alerts by querying Kubernetes, Prometheus, logs and other sources with an LLM and summarises likely root causes. CNCF sandbox project started at Robusta.
Pricing and hosting
- Pricing
- Free
- Open source
- Yes (Apache-2.0)
- Self-hosted
- Yes
- Launched
- 2024
- Status
- Active
- Website
- holmesgpt.dev
Plans and prices change; check the tool's own site before you decide.
What it does
Incident management checklist
1 of 10 confirmed in HolmesGPT's own documentation. The same checklist is used for every tool in Incident management.
- Declare from chat No
- Incident channels No
- Incident roles No
- Severity levels No
- Timeline Not confirmed
- Workflows Yes
- Status page updates No
- Postmortems Not confirmed
- Follow-up actions Not confirmed
- Paging No
Open-source investigation agent, not an incident manager: no incident record, channels, roles, status page or paging. Workflows = automatic investigations (operator mode, alert integrations).
Checked against: github.com, holmesgpt.dev, holmesgpt.dev (2)
Alternatives to HolmesGPT
Other incident management tools. See all 20.
-
AWS Systems Manager Incident Manager AWS service for response plans, on-call contacts and escalation plans, engagement by SMS and voice, and runbook automation through Systems Manager. -
Cleric AI SRE agent that investigates production alerts on its own by querying observability and infrastructure tools, then reports findings and likely causes to the on-call engineer. -
Cutover Runbook platform for large organisations that turns disaster recovery, cloud migration, release and major-incident procedures into shared, partly automated task lists. -
Dispatch Netflix's open source incident management tool that creates Slack channels, documents and tickets for each incident and tracks roles, tasks and timelines. -
Event-Driven Ansible Rulebooks that listen to events from monitoring tools, webhooks or Kafka and trigger Ansible playbooks, used for automatic remediation. Open source core, supported in Red Hat Ansible Automation Platform. -
FireHydrant Incident management platform with runbooks, Slack and Teams incident channels, retrospectives and status pages. Bought Blameless in 2024 and agreed to join Freshworks in December 2025. -
Harness AI SRE Incident response module in the Harness platform that gathers alerts and change events, runs response workflows and uses AI agents to suggest causes. -
Incident Bot Self-hosted Slack or Matrix bot for incident response: opens channels, tracks severity, status and roles, posts a digest and builds postmortem documents. -
incident.io Incident management run from Slack or Microsoft Teams: declares incidents, assigns roles, keeps timelines, automates workflows and drafts postmortems, with status pages and on-call. -
K8sGPT CLI and operator that scans Kubernetes clusters for problems and explains them in plain language using an LLM backend of your choice. CNCF sandbox project. -
NeuBird AI SRE product that investigates alerts and incidents by querying your existing telemetry and cloud tools, and reports probable causes; priced per investigation rather than per token. -
Resolve AI AI SRE product whose agents investigate alerts and incidents across code, infrastructure and telemetry and take on routine operational tasks for on-call engineers.