feat(services): env doctor — surface empty vars, bad refs, unreadable secrets

GET /services/:id/env-check (read scope) reports, without exposing values:
- unresolved: ${{service.KEY}} refs that don't match a sibling service/env
- empty: variables with blank values
- unreadable: secrets whose ciphertext fails decryption (key rotation)

install.sh now generates SECRETS_KEY explicitly (was falling back to
JWT_SECRET — a rotation would orphan ciphertext). .env.example documents it.

Surfaces: CLI `services env-check`, MCP containr_services_env_check
(90 tools), OpenAPI path, and a warning banner in the service Variables
panel listing each issue.

Verified live: injected a bogus ${{nosuchservice.PORT}} var plus an
unreadable ciphertext row; the endpoint flagged both, ok=false.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

.env.example

@@ -39,6 +39,10 @@ REDIS_URL=redis://:your_secure_redis_password@localhost:6379/0
 # Application Configuration
 # In production this must be a strong value with at least 32 characters.
 JWT_SECRET=SuperSecretJWTKey123456789!@#$%^&*()
+# Encrypts secret env vars at rest. Defaults to JWT_SECRET when unset;
+# setting it explicitly survives a JWT_SECRET rotation.
+SECRETS_KEY=
+
 # Shared secret for node-agent registration/heartbeat auth.
 CONTAINR_AGENT_AUTH_TOKEN=replace_with_strong_agent_secret
 # Optional rotation list (comma-separated). When set, this takes precedence.

app/backend/internal/api/routes.go

@@ -279,6 +279,7 @@ func SetupRoutes(router *gin.Engine, db *database.DB, redis *database.Redis, cfg
 			read.GET("/services/:id/deployments", handleGetDeployments)
 			read.GET("/services/:id/domains", handleListServiceDomains)
 			read.GET("/services/:id/domains/check", handleCheckServiceDomains)
+			read.GET("/services/:id/env-check", handleServiceEnvCheck)
 			read.GET("/deployments", handleGetRecentDeployments)
 			read.GET("/deployments/:id", handleGetDeployment)
 

app/backend/internal/api/runtime.go

@@ -638,3 +638,112 @@ func runtimeScaleService(c *gin.Context, serviceID string, replicas int) error {
 	_, _ = db.Exec(`UPDATE services SET replicas = $1, updated_at = $2 WHERE id = $3`, replicas, time.Now(), service.ID)
 	return nil
 }
+
+// handleServiceEnvCheck reports environment problems without exposing
+// values: empty vars, ${{service.KEY}} refs that fail to resolve, and
+// secrets whose ciphertext can't be decrypted (key rotation).
+func handleServiceEnvCheck(c *gin.Context) {
+	dbVal, exists := c.Get("db")
+	if !exists {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": "Database connection not available"})
+		return
+	}
+	db := dbVal.(*database.DB)
+	service, ok := loadReadableService(c, db)
+	if !ok {
+		return
+	}
+
+	rows, err := db.Query(
+		`SELECT key, value, COALESCE(is_secret, false) FROM environment_variables WHERE service_id = $1`,
+		service.ID,
+	)
+	if err != nil {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	defer rows.Close()
+
+	var unresolved, empty, unreadable []string
+	var raw []struct{ key, val string }
+	for rows.Next() {
+		var k, v string
+		var secret bool
+		if err := rows.Scan(&k, &v, &secret); err != nil {
+			continue
+		}
+		decrypted := secrets.Decrypt(v)
+		if secret && secrets.IsEncrypted(v) && decrypted == v {
+			// Ciphertext failed to decrypt — Decrypt returns the input.
+			unreadable = append(unreadable, k)
+			continue
+		}
+		if strings.TrimSpace(decrypted) == "" {
+			empty = append(empty, k)
+		}
+		raw = append(raw, struct{ key, val string }{k, decrypted})
+	}
+
+	// Project siblings for ${{name.KEY}} resolution.
+	siblingIDs := map[string]string{}
+	svcRows, err := db.Query(`SELECT id, name FROM services WHERE project_id = $1`, service.ProjectID)
+	if err == nil {
+		for svcRows.Next() {
+			var id, name string
+			if err := svcRows.Scan(&id, &name); err == nil {
+				siblingIDs[strings.ToLower(name)] = id
+			}
+		}
+		svcRows.Close()
+	}
+
+	resolvable := func(serviceName, key string) bool {
+		id, ok := siblingIDs[strings.ToLower(serviceName)]
+		if !ok {
+			return false
+		}
+		if strings.EqualFold(key, "HOST") {
+			return true
+		}
+		if strings.EqualFold(key, "PORT") {
+			var port int
+			if err := db.QueryRow(`SELECT COALESCE(port, 0) FROM services WHERE id = $1`, id).Scan(&port); err == nil {
+				return port > 0
+			}
+			return false
+		}
+		var n int
+		if err := db.QueryRow(
+			`SELECT COUNT(*) FROM environment_variables WHERE service_id = $1 AND key = $2`,
+			id, key,
+		).Scan(&n); err == nil {
+			return n > 0
+		}
+		return false
+	}
+
+	for _, kv := range raw {
+		for _, match := range refVarPattern.FindAllStringSubmatch(kv.val, -1) {
+			if len(match) == 3 && !resolvable(match[1], match[2]) {
+				unresolved = append(unresolved, fmt.Sprintf("%s → %s", kv.key, match[0]))
+			}
+		}
+	}
+
+	if unresolved == nil {
+		unresolved = []string{}
+	}
+	if empty == nil {
+		empty = []string{}
+	}
+	if unreadable == nil {
+		unreadable = []string{}
+	}
+	okResult := len(unresolved) == 0 && len(empty) == 0 && len(unreadable) == 0
+	c.JSON(http.StatusOK, gin.H{
+		"ok":         okResult,
+		"unresolved": unresolved,
+		"empty":      empty,
+		"unreadable": unreadable,
+	})
+}

app/backend/internal/cli/commands/services.go

@@ -530,5 +530,38 @@ func init() {
 
 	domainsAddCmd.Flags().Bool("default", false, "set as the default domain")
 	servicesDomainsCmd.AddCommand(domainsListCmd, domainsAddCmd, domainsRemoveCmd, domainsDefaultCmd, domainsCheckCmd)
-	ServicesCmd.AddCommand(servicesListCmd, servicesGetCmd, servicesCreateCmd, servicesUpdateCmd, servicesDeleteCmd, servicesDomainsCmd, servicesCloneCmd, servicesMoveCmd)
+	ServicesCmd.AddCommand(servicesListCmd, servicesGetCmd, servicesCreateCmd, servicesUpdateCmd, servicesDeleteCmd, servicesDomainsCmd, servicesCloneCmd, servicesMoveCmd, servicesEnvCheckCmd)
+}
+
+var servicesEnvCheckCmd = &cobra.Command{
+	Use:   "env-check <id>",
+	Short: "Diagnose service environment: empty vars, unresolved ${{refs}}",
+	Args:  cobra.ExactArgs(1),
+	RunE: func(cmd *cobra.Command, args []string) error {
+		c, err := client()
+		if err != nil {
+			return err
+		}
+		data, err := c.Do("GET", "/services/"+args[0]+"/env-check", nil)
+		if err != nil {
+			return err
+		}
+		if JSONMode() {
+			PrintRaw(data)
+			return nil
+		}
+		m, _ := unwrapObject(data)
+		if b, ok := m["ok"].(bool); ok && b {
+			fmt.Println("env check: ok")
+			return nil
+		}
+		for _, group := range []string{"unresolved", "empty", "unreadable"} {
+			if list, ok := m[group].([]interface{}); ok {
+				for _, item := range list {
+					fmt.Printf("%s: %v\n", group, item)
+				}
+			}
+		}
+		return &APIError{Message: "env check failed", ExitCode: ExitError}
+	},
 }

app/backend/internal/mcpserver/server.go

@@ -60,6 +60,7 @@ func init() {
 	register(spec{"containr_services_restart", "Restart a service", "POST", "/services/{id}/restart", []string{"id"}, nil, false})
 	register(spec{"containr_services_sleep", "Put a service to sleep immediately (scale to zero)", "POST", "/services/{id}/sleep", []string{"id"}, nil, false})
 	register(spec{"containr_services_wake", "Wake a sleeping service", "POST", "/services/{id}/wake", []string{"id"}, nil, false})
+	register(spec{"containr_services_env_check", "Diagnose service env: returns ok plus unresolved ${{refs}}, empty vars, and undecryptable secrets", "GET", "/services/{id}/env-check", []string{"id"}, nil, false})
 	register(spec{"containr_services_redeploy", "Rebuild and redeploy a service", "POST", "/services/{id}/redeploy", []string{"id"}, nil, false})
 	register(spec{"containr_services_clone", "Clone a service (config, volumes, domains, variables) into the same or another project. Args: id, name, project_id, environment.", "POST", "/services/{id}/clone", []string{"id"}, nil, false})
 	register(spec{"containr_services_move", "Move a service to another project the caller owns. Args: id, project_id (required).", "POST", "/services/{id}/move", []string{"id"}, []string{"project_id"}, false})

app/frontend/src/features/service/pages/ServiceDetailPage.tsx

@@ -36,6 +36,7 @@ import {
   sleepService,
   wakeService,
   getServiceRuntime,
+  getServiceEnvCheck,
   type CronJobEntity,
   type ServiceVolume,
   type ServiceDomain,
@@ -189,6 +190,12 @@ export function ServiceDetailPage() {
     enabled: Boolean(serviceId) && !isDemoMode && activeSection === 'variables',
   });
 
+  const envCheckQuery = useQuery({
+    queryKey: ['service-env-check', serviceId],
+    queryFn: () => getServiceEnvCheck(serviceId),
+    enabled: Boolean(serviceId) && !isDemoMode && activeSection === 'variables',
+  });
+
   const saveVariablesMutation = useMutation({
     mutationFn: (rows: VariableDraft[]) =>
       updateServiceVariables(
@@ -1315,6 +1322,23 @@ export function ServiceDetailPage() {
               </div>
             </div>
 
+            {envCheckQuery.data && !envCheckQuery.data.ok && (
+              <div className="mb-4 rounded-[var(--radius-md)] border border-[var(--warning)]/40 bg-[var(--warning-soft)] p-4">
+                <p className="text-xs font-semibold text-[var(--warning)] mb-2">Environment issues detected</p>
+                <ul className="space-y-1 text-xs text-[var(--text-secondary)]">
+                  {(envCheckQuery.data.unresolved ?? []).map((u) => (
+                    <li key={u} className="mono">Unresolved reference: {u}</li>
+                  ))}
+                  {(envCheckQuery.data.empty ?? []).map((k) => (
+                    <li key={k} className="mono">Empty value: {k}</li>
+                  ))}
+                  {(envCheckQuery.data.unreadable ?? []).map((k) => (
+                    <li key={k} className="mono">Cannot decrypt secret (key rotated?): {k}</li>
+                  ))}
+                </ul>
+              </div>
+            )}
+
             {bulkOpen && (
               <div className="mb-4 panel-soft p-4">
                 <label className="block text-xs font-medium uppercase tracking-wider text-[var(--text-muted)] mb-2">

app/frontend/src/generated/api-types.ts

@@ -1010,6 +1010,52 @@ export interface paths {
         patch?: never;
         trace?: never;
     };
+    "/services/{id}/env-check": {
+        parameters: {
+            query?: never;
+            header?: never;
+            path?: never;
+            cookie?: never;
+        };
+        /**
+         * Environment diagnostics
+         * @description Reports empty variables, unresolved ${{service.KEY}} references, and secrets whose stored ciphertext fails to decrypt. Values are never returned.
+         */
+        get: {
+            parameters: {
+                query?: never;
+                header?: never;
+                path: {
+                    id: string;
+                };
+                cookie?: never;
+            };
+            requestBody?: never;
+            responses: {
+                /** @description Diagnostic result */
+                200: {
+                    headers: {
+                        [name: string]: unknown;
+                    };
+                    content: {
+                        "application/json": {
+                            ok?: boolean;
+                            unresolved?: string[];
+                            empty?: string[];
+                            unreadable?: string[];
+                        };
+                    };
+                };
+            };
+        };
+        put?: never;
+        post?: never;
+        delete?: never;
+        options?: never;
+        head?: never;
+        patch?: never;
+        trace?: never;
+    };
     "/internal/wake/{id}": {
         parameters: {
             query?: never;

app/frontend/src/generated/docs-snapshot.ts

@@ -40,7 +40,7 @@ export const bundledDocs: BundledDoc[] = [
     "path": "guides/CLI.md",
     "section": "guides",
     "title": "Containr CLI",
-    "markdown": "# Containr CLI\n\nThe CLI controls every part of a Containr instance — the same API surface\nthe web UI uses. Build it with `scripts/build-cli.sh` or\n`go build ./cmd/cli` from `app/backend`.\n\n## Authentication\n\nCreate a personal access token in the web UI under\n**Settings → Personal Access Tokens** (or `containr tokens create`), then:\n\n```bash\ncontainr auth login cnp_...                      # verify + store\ncontainr auth login cnp_... --url https://containr.example.com\ncontainr auth login cnp_... --name production    # named profile\ncontainr auth status                             # show active profile + identity\ncontainr auth logout\n```\n\nResolution order for every command: `--token`/`--api-url` flags >\n`CONTAINR_TOKEN`/`CONTAINR_API_URL` env > active profile >\n`http://localhost:8080/api/v1`. The active profile is `--profile` >\n`CONTAINR_PROFILE` > `current_profile` in `~/.containr.yaml`.\n\n## Output and scripting\n\n- `--json` prints the raw API response — safe to pipe and parse.\n- `--yes` / `-y` skips confirmation on destructive commands (delete,\n  revoke, restore, failover). Non-interactive shells refuse destructive\n  commands without it.\n- Exit codes are stable: `0` ok, `1` error, `2` auth/permission,\n  `3` not found, `4` validation, `5` unreachable.\n\n## Commands\n\n```\nauth login|logout|status         token auth and profiles\ntokens list|create|revoke        personal access tokens\nprojects list|get|create|delete\nservices list|get|create|update|delete|start|stop|restart|redeploy\nservices clone|move              duplicate config or re-parent a service\nservices domains list|add|remove|default|check   multi-domain + DNS preflight\nregistries list|add|update|remove  private-image pull credentials\nvolumes list|delete              docker volume inventory + cleanup (admin)\ndeploy <service-id>              trigger a deployment (--no-cache, --commit-hash, --branch)\ndeployments list|get|logs|rollback|cancel\nlogs <service-id> [--follow]     runtime logs (SSE follow)\nexec <service-id> -- <cmd>       one-off container command\nvariables list|set|unset         env vars (secrets masked; --redeploy applies them)\ndatabases list|get|create|update|delete|action|backup|restore|download-backup|import-backup\n                               import-backup <id> <file.tar.gz|-> uploads an archive as a restore\n                               point; ships offsite too when a backup target is assigned\n                               create --public exposes the port on all interfaces;\n                               update <id> --public/--public=false toggles it (container recreates, data persists)\ndatabases register <name>      register an external database (--type --host --port --database --username --password --ssl; probed before storing)\ndatabases test-connection      probe a connection — by id, or ad-hoc flags\nbackup-targets list|add|update|remove|test\n                               S3-compatible archive destinations; add/update probe the bucket\n                               first, empty --access-key/--secret-key on update keep stored keys;\n                               update <db-id> --backup-target assigns a target (\"\" clears it)\ncron list|get|create|delete|trigger|executions\ntemplates list|get|create|delete|deploy|plan|import-compose|deploy-graph\n                               deploy --name --var K=V; plan dry-resolves expressions;\n                               import-compose <compose.yml|-> converts a compose file into\n                               a v2 graph config; deploy-graph <project-id> <config.json|->\n                               deploys an ad-hoc service graph\nnodes list|get|delete|tokens     node agents + onboarding tokens (admin)\nscaling status|policies|services|scale\nha status|alerts|health|policies|enable|disable|failover\nsecurity scan|vulnerabilities|metrics\ngateway services|keys|analytics  (admin)\nadmin overview|users|settings|audit-logs\nnotifications list|read\nup                               deploy the current directory\n```\n\n## containr up\n\nRun inside a git checkout: uses `remote.origin.url` + the current branch\nas the build source, creates a service named after the directory when none\nexists, then triggers a deployment.\n\n```bash\ncontainr up --project <id>\ncontainr up --project <id> --image ghcr.io/example/app:latest   # image, no build\n```\n\nSet `project_id` on the active profile to omit `--project` (stored in\n`~/.containr.yaml` under `profiles.<name>.project_id`).\n\n## Service volumes\n\n`services update` manages mounts; flags replace the whole mount list:\n\n```bash\ncontainr services update <id> --volume data:/data --bind /srv/cfg:/etc/app:ro\ncontainr services update <id> --clear-volumes\n```\n\n`type` is `volume` (docker named volume, auto-created) or `bind` (host\npath). `target` must be absolute. Mounts apply on the next deploy or\nredeploy — `containr deploy <id>` to apply immediately.\n\n`voLine truncated
+    "markdown": "# Containr CLI\n\nThe CLI controls every part of a Containr instance — the same API surface\nthe web UI uses. Build it with `scripts/build-cli.sh` or\n`go build ./cmd/cli` from `app/backend`.\n\n## Authentication\n\nCreate a personal access token in the web UI under\n**Settings → Personal Access Tokens** (or `containr tokens create`), then:\n\n```bash\ncontainr auth login cnp_...                      # verify + store\ncontainr auth login cnp_... --url https://containr.example.com\ncontainr auth login cnp_... --name production    # named profile\ncontainr auth status                             # show active profile + identity\ncontainr auth logout\n```\n\nResolution order for every command: `--token`/`--api-url` flags >\n`CONTAINR_TOKEN`/`CONTAINR_API_URL` env > active profile >\n`http://localhost:8080/api/v1`. The active profile is `--profile` >\n`CONTAINR_PROFILE` > `current_profile` in `~/.containr.yaml`.\n\n## Output and scripting\n\n- `--json` prints the raw API response — safe to pipe and parse.\n- `--yes` / `-y` skips confirmation on destructive commands (delete,\n  revoke, restore, failover). Non-interactive shells refuse destructive\n  commands without it.\n- Exit codes are stable: `0` ok, `1` error, `2` auth/permission,\n  `3` not found, `4` validation, `5` unreachable.\n\n## Commands\n\n```\nauth login|logout|status         token auth and profiles\ntokens list|create|revoke        personal access tokens\nprojects list|get|create|delete\nservices list|get|create|update|delete|start|stop|restart|redeploy\nservices clone|move              duplicate config or re-parent a service\nservices domains list|add|remove|default|check   multi-domain + DNS preflight\nregistries list|add|update|remove  private-image pull credentials\nvolumes list|delete              docker volume inventory + cleanup (admin)\ndeploy <service-id>              trigger a deployment (--no-cache, --commit-hash, --branch)\ndeployments list|get|logs|rollback|cancel\nlogs <service-id> [--follow]     runtime logs (SSE follow)\nexec <service-id> -- <cmd>       one-off container command\nvariables list|set|unset         env vars (secrets masked; --redeploy applies them)\ndatabases list|get|create|update|delete|action|backup|restore|download-backup|import-backup\n                               import-backup <id> <file.tar.gz|-> uploads an archive as a restore\n                               point; ships offsite too when a backup target is assigned\n                               create --public exposes the port on all interfaces;\n                               update <id> --public/--public=false toggles it (container recreates, data persists)\ndatabases register <name>      register an external database (--type --host --port --database --username --password --ssl; probed before storing)\ndatabases test-connection      probe a connection — by id, or ad-hoc flags\nbackup-targets list|add|update|remove|test\n                               S3-compatible archive destinations; add/update probe the bucket\n                               first, empty --access-key/--secret-key on update keep stored keys;\n                               update <db-id> --backup-target assigns a target (\"\" clears it)\ncron list|get|create|delete|trigger|executions\ntemplates list|get|create|delete|deploy|plan|import-compose|deploy-graph\n                               deploy --name --var K=V; plan dry-resolves expressions;\n                               import-compose <compose.yml|-> converts a compose file into\n                               a v2 graph config; deploy-graph <project-id> <config.json|->\n                               deploys an ad-hoc service graph\nnodes list|get|delete|tokens     node agents + onboarding tokens (admin)\nscaling status|policies|services|scale\nha status|alerts|health|policies|enable|disable|failover\nsecurity scan|vulnerabilities|metrics\ngateway services|keys|analytics  (admin)\nadmin overview|users|settings|audit-logs\nnotifications list|read\nup                               deploy the current directory\n```\n\n## containr up\n\nRun inside a git checkout: uses `remote.origin.url` + the current branch\nas the build source, creates a service named after the directory when none\nexists, then triggers a deployment.\n\n```bash\ncontainr up --project <id>\ncontainr up --project <id> --image ghcr.io/example/app:latest   # image, no build\n```\n\nSet `project_id` on the active profile to omit `--project` (stored in\n`~/.containr.yaml` under `profiles.<name>.project_id`).\n\n## Service volumes\n\n`services update` manages mounts; flags replace the whole mount list:\n\n```bash\ncontainr services update <id> --volume data:/data --bind /srv/cfg:/etc/app:ro\ncontainr services update <id> --clear-volumes\n```\n\n`type` is `volume` (docker named volume, auto-created) or `bind` (host\npath). `target` must be absolute. Mounts apply on the next deploy or\nredeploy — `containr deploy <id>` to apply immediately.\n\n`voLine truncated
   },
   {
     "path": "guides/CLOUDFLARE_SETUP.md",

app/frontend/src/lib/api-client.ts

@@ -1304,6 +1304,17 @@ export const restartService = (serviceId: string) => serviceAction(serviceId, 'r
 export const sleepService = (serviceId: string) => serviceAction(serviceId, 'sleep');
 export const wakeService = (serviceId: string) => serviceAction(serviceId, 'wake');
 
+export type EnvCheckResult = {
+  ok: boolean;
+  unresolved?: string[];
+  empty?: string[];
+  unreadable?: string[];
+};
+
+export async function getServiceEnvCheck(serviceId: string): Promise<EnvCheckResult> {
+  return requestJson<EnvCheckResult>(`/services/${serviceId}/env-check`);
+}
+
 export async function redeployService(serviceId: string): Promise<ServiceRuntime> {
   const payload = await requestJson<{ runtime?: ServiceRuntime }>(`/services/${serviceId}/redeploy`, {
     method: 'POST',

docs/api/openapi.yaml

@@ -685,6 +685,41 @@ paths:
         '202':
           description: Wake initiated
 
+  /services/{id}/env-check:
+    get:
+      tags:
+        - Services
+      summary: Environment diagnostics
+      description: Reports empty variables, unresolved ${{service.KEY}} references, and secrets whose stored ciphertext fails to decrypt. Values are never returned.
+      parameters:
+        - name: id
+          in: path
+          required: true
+          schema:
+            type: string
+      responses:
+        '200':
+          description: Diagnostic result
+          content:
+            application/json:
+              schema:
+                type: object
+                properties:
+                  ok:
+                    type: boolean
+                  unresolved:
+                    type: array
+                    items:
+                      type: string
+                  empty:
+                    type: array
+                    items:
+                      type: string
+                  unreadable:
+                    type: array
+                    items:
+                      type: string
+
   /internal/wake/{id}:
     get:
       tags:

docs/guides/CLI.md

@@ -168,6 +168,7 @@ containr services update <id> --sleep --sleep-idle 15   # enable, 15 min idle ti
 containr services update <id> --sleep=false            # disable
 containr services sleep <id>                           # sleep now
 containr services wake <id>                            # resume now
+containr services env-check <id>                      # flag empty vars + unresolved ${{refs}}
 ```
 
 Idleness is measured as flat container network I/O over the configured

install.sh

@@ -73,6 +73,7 @@ else
     PG_PASS="$(gen)"; REDIS_PASS="$(gen)"
     setvar ENVIRONMENT "production"
     setvar JWT_SECRET "$(gen)"
+    setvar SECRETS_KEY "$(gen)"
     setvar BETTER_AUTH_SECRET "$(gen)"
     setvar BETTER_AUTH_INTERNAL_TOKEN "$(gen)"
     setvar CONTAINR_AGENT_AUTH_TOKEN "$(gen)"

llms.txt

@@ -25,7 +25,7 @@ Base URL: `https://<host>/api/v1`. OpenAPI: `docs/api/openapi.yaml`.
 - **Public DB port**: managed databases bind `127.0.0.1` by default; `public_port:true` on create or `PUT /databases/{id} {public_port}` binds `0.0.0.0`. Toggling recreates the container — the data volume persists.
 - **Backup targets**: S3-compatible destinations (`/backup-targets` CRUD + `POST /backup-targets/{id}/test`). Endpoints/buckets are probed before storing; keys encrypted at rest (`has_credentials` only). Assign via `PUT /databases/{id} {backup_target_id}` — empty string clears; external databases reject assignment. Managed backups stay local; a copy ships to `[prefix/]databases/<db-id>/<archive>` and rows expose `remote:true`. Restore prefers the local archive and falls back to the remote object. Deleting a target is refused (409) while databases reference it.
 - **Backup import**: `POST /databases/{id}/backups/import` — upload a `.tar.gz` (multipart `file` or raw `application/gzip` body, ≤2 GiB) to create a restore point; also ships offsite when a target is assigned. CLI: `databases import-backup <id> <file|->`.
-- **Sleep mode**: `PUT /services/{id} {sleep_enabled, sleep_idle_minutes}` scales a service to zero after N idle minutes (network-byte deltas, 30s sweep). Status becomes `sleeping`. A busybox wake container then holds the service's Traefik host rule at priority 100, serves a reloading "waking" page, and hits `GET /api/v1/internal/wake/{id}` (unauthenticated, idempotent) which reconciles the service back. `POST /services/{id}/sleep|wake` for manual control. Requires the API reachable from containers (containerized deploy or `HOST=0.0.0.0`).
+- **Sleep mode**: `PUT /services/{id} {sleep_enabled, sleep_idle_minutes}` scales a service to zero after N idle minutes (network-byte deltas, 30s sweep). Status becomes `sleeping`. A busybox wake container then holds the service's Traefik host rule at priority 100, serves a reloading "waking" page, and hits `GET /api/v1/internal/wake/{id}` (unauthenticated, idempotent) which reconciles the service back. `POST /services/{id}/sleep|wake` for manual control; `GET /services/{id}/env-check` reports empty vars, unresolved `${{svc.KEY}}` refs, and undecryptable secrets (values never exposed). Requires the API reachable from containers (containerized deploy or `HOST=0.0.0.0`).
 - **Template graphs (v2)**: a template `config` with `services[]` deploys a multi-service stack. Members declare `key`, `type` (`web|worker|cron|database`), `runtime`, `environment`, `depends_on`. Env values support expressions: `{{secret}}`, `{{secret(N)}}`, `{{random:N}}`, `{{VAR}}`, `{{VAR:-default}}`, `{{service.<key>.<prop>}}` (props: `host`, `port`, `url`, `user`, `password`, `database`). `database` members become managed databases attached to the project network under their key — siblings resolve them by name. `POST /templates/{id}/plan` dry-resolves; `POST /templates/deploy` deploys an inline graph; `POST /templates/import/compose` converts docker-compose YAML (known db images → managed databases). Service↔service refs `${{svc.KEY}}` also resolve at runtime (HOST/PORT or env vars).
 - **Identifiers**: all ids are UUIDs. Tokens: `cnp_` user PATs, `cagt_` agent onboarding tokens.