Cloud
Cloud docs describe how TomoriBot’s production infrastructure runs on Azure: deployment targets, log/metric ingestion, and the operational rules around them. The AWS and GCP deployment code is retained as legacy rollback and self-hosting reference; TomoriBot’s managed GCP project was retired after the Azure cutover on 2026-07-20. Provider-agnostic hosting on your own machines lives under Self-Hosting instead.
A small dedicated GCP project is retained only as an outbound AI provider. It does not host TomoriBot infrastructure.
azure-production-deployment.md— OIDC/Run Command deployment, database-role bootstrap, host lockdown, and production operating rulesazure-application-logs.md— shipping structured error logs from an Azure VM into Log Analytics and Grafanaazure-vertex-auth.md— keyless authentication from the Azure VM to Google Vertex AI
Two procedures act on one specific deployment rather than describing the architecture, so they live in the hidden wiki instead of here: Terraform State Recovery and Production Data Inspection.