feat: implement proto-gen platform

- Config loader with env var expansion and validation
- SQLite database with tasks and toolchain_versions tables
- Gitea REST API client for repository and branch operations
- K8s client for Job lifecycle management (with noop fallback for local dev)
- REST API: tasks CRUD, retry, plugins, version management
- Worker/scheduler with polling, concurrency control, and job tracking
- Static frontend: generate, tasks, settings pages
- Toolchain images: Go and Node.js Dockerfiles with entrypoint scripts
- Helm chart with deployment, service, configmap, ingress, PVC
- Multi-stage Dockerfile for service
- Unit tests for config, database, gitclient, k8s, api, worker
- E2E tests for idempotency, version switch, failure recovery

deploy/helm/proto-gen/values.yaml

@@ -0,0 +1,54 @@
+replicaCount: 1
+
+image:
+  repository: registry.example.com/proto-gen
+  tag: "latest"
+  pullPolicy: IfNotPresent
+
+git:
+  baseUrl: "https://git.example.com"
+  tokenSecret: "proto-gen-git-token"
+
+cluster:
+  kubeconfigSecret: "proto-gen-kubeconfig"
+  namespace: proto-gen-jobs
+  jobTimeout: 600
+  maxConcurrent: 2
+
+toolchainRegistry: "registry.example.com/proto-gen"
+
+targetRepoTemplate: "{proto_repo}-gen-{lang}"
+
+plugins:
+  - language: go
+    image: "proto-gen-go:v1.34.2"
+  - language: nodejs
+    image: "proto-gen-ts:v0.9.0"
+
+service:
+  type: ClusterIP
+  port: 8080
+
+ingress:
+  enabled: false
+  className: ""
+  annotations: {}
+  hosts:
+    - host: gen.example.com
+      paths:
+        - path: /
+          pathType: Prefix
+  tls: []
+
+persistence:
+  enabled: true
+  storageClass: ""
+  size: 1Gi
+
+resources: {}
+
+nodeSelector: {}
+
+tolerations: []
+
+affinity: {}