Merge branch 'main' into mark-swift-data-grid-header-toolbar

Author: TilmanHaupt

.changeset/1781603633-policy-json-migration.md

@@ -0,0 +1,18 @@
+---
+"@cobaltcore-dev/aurora": minor
+"@cobaltcore-dev/dashboard": minor
+---
+
+Migrate policy files from YAML to JSON and unify storage policies
+
+- Convert all policy files from YAML to JSON format (compute, image, networking, storage)
+- Unify Swift and Ceph policies into single storage.json with consistent Swift terminology
+- Add startup validation for engine configuration in createPermissionRouter
+- Update router definitions to use .json filenames instead of .yaml
+
+Benefits:
+- Better tooling support (schema validation, editor autocomplete)
+- Consistent naming: storage:containers:*, storage:objects:*, storage:folders:*
+- Backend-agnostic API (UI doesn't distinguish Swift vs Ceph)
+- Fewer files to maintain (4 instead of 6 policy files)
+- Errors caught at startup instead of runtime

.changeset/afraid-worms-dress.md

@@ -0,0 +1,5 @@
+---
+"@cobaltcore-dev/aurora": patch
+---
+
+Compute, network, and service overview routes now redirect to the project overview

.changeset/brave-eagles-rescue.md

@@ -0,0 +1,22 @@
+---
+"@cobaltcore-dev/aurora": minor
+---
+
+feat(metrics): add Prometheus metrics for infrastructure monitoring
+
+Added comprehensive Prometheus metrics collection for HTTP requests, including:
+
+- `aurora_requests_total`: Counter tracking total HTTP requests with labels for status_code, method, route, endpoint_type, and project_id
+- `aurora_request_duration_seconds`: Histogram measuring request latency with the same label dimensions
+- `aurora_exceptions_total`: Counter tracking unhandled exceptions by exception type
+
+Key features:
+- Intelligent route normalization to control cardinality (tRPC procedures, static assets, SPA routes, Vite dev paths)
+- Per-project visibility via project_id label extracted from URLs and query parameters
+- tRPC batch request support (comma-separated procedures in route label)
+- Excludes /metrics endpoint from collection to prevent recursion
+- Debug logging for skipped metrics to aid troubleshooting
+
+New endpoint: `GET /metrics` exposes metrics in Prometheus text format for scraping.
+
+Note: project_id label creates one time series per unique project. Monitor Prometheus memory usage in production deployments with thousands of projects.