The data foundation for modern automation

Infrahub on GitHub · Sandbox · Documentation · Discord

OpsMill builds Infrahub, the open-source data and automation platform for network, data center, and cloud infrastructure. Infrahub offers teams a graph-based, schema-first source of truth — purpose-built to be the validated data layer that automation tools and AI agents both depend on.

Most large-scale automation efforts do not fail because of automation logic. They fail because of the data underneath it.

Infrastructure data is typically fragmented across spreadsheets, IPAMs, CMDBs, and YAML files — without a consistent schema, version control, or reliable way to model relationships. This leads to:

• Fragile automation that is difficult to scale
• Risky changes with limited validation
• Significant operational overhead maintaining data consistency

Infrahub treats infrastructure data the way modern software treats source code: modeled, versioned, validated, and queryable.

• Multi-vendor configuration management with validation and drift detection
• Brownfield automation alongside existing tools

• Model services and automate the full lifecycle (provision → upgrade → decommission)
• Build self-service catalogs backed by infrastructure data

• Deploy repeatable environments (e.g. GPU fabrics, multi-site architectures)
• Reuse schemas and blueprints across environments

• Model relationships as a knowledge graph
• Enable AI systems to reason over infrastructure dependencies

'Custom-built' platforms can match any of these capabilities — but at the cost of years of build time and ongoing maintenance. For a deeper side-by-side comparison, see How Infrahub compares.

Infrahub does not replace orchestration or monitoring systems. It provides the validated, intent-rich data those systems depend on.

OpsMill maintains a set of officially supported open-source projects that extend Infrahub across four areas:

• Developer tooling — a Python SDK and infrahubctl CLI, a VS Code extension with schema validation and visualization, and a reusable schema library to accelerate modeling.
• Production operations — Helm charts for Kubernetes deployment, plus backup and restore tooling.
• Integrations — an Ansible Collection, a Nornir plugin, and Infrahub Sync for federating data from Netbox, Nautobot, IP Fabric, and other systems.
• AI agents and assistants — an MCP server, Infrahub Skills for Claude, Cursor, and other AI coding assistants, and a Speckit preset for spec-driven AI development.

See github.com/opsmill for the full list of projects.

• Website — opsmill.com
• Documentation — docs.infrahub.app
• Blog — opsmill.com/blog
• Newsletter — Ottermatic
• Discord — discord.gg/opsmill

Infrahub is available under Apache 2.0 as a free Community edition, and as Infrahub Enterprise for teams running it at production scale.