Skip to content

Platform
Platform overviewThe operating model, end to end ArchitectureTopology, boundaries and blast radius Workflow orchestrationScripts, AI goals and approvals in one DAG GitOps configurationReviewable, promotable platform intent Governance and AAADefault-deny, caller-scoped decisions AI and MCPApproved models, tools, context and limits Knowledge and contextThe policies and guardrails you bind to a step Runtime and executionDocker and Kubernetes placement Evidence and monitoringDurable proof attached to the run
Use cases
All use casesTen workflows shown as governed runs Production incidentDiagnose and roll back behind an approval gate Release preparationScope, gates, notes and change request Hotfix to productionOne trail from complaint to emergency change Security scan triageFindings to scoped, tracked fixes
Why NopsAI
Integrations
Security
Resources
All resourcesDefinitions, comparisons and evaluation checklists AI agent governanceBoundaries that survive production MCP governanceFour decisions before a tool is available MCP securityControls when a server can reach production Self-hosted platformsThe evaluation checklist a security review uses
Pricing

Docs Book a demo

Automation

How runs start: manually, from a Git event, on a schedule, from an external system, or from a parent pipeline.

Automation

ConceptHow runs startThe six entry points into a run, and what each one controls about ownership and scope.ReferenceGit trigger manifestsThe repository-side `.nopsai.yaml` that decides which events start which pipelines.How-toGitHub App and installationsOne GitHub App per NopsAI installation, installed on as many organizations and personal accounts as you need.How-toGit webhook sourcesManaged ingress for GitLab, Bitbucket, Gitea, and generic providers.ReferenceExternal API triggersLet another system start a run, with explicit callers, payload validation, mapping, and rate limits.ReferenceSchedulesCron and one-time schedules, their ownership, and the variables they supply.