Add per-deployment GraphQL retry in release phase

The release phase runs 8+ deployments in parallel via runNamedParallel,
which can overwhelm the Dagger engine and cause transient GraphQL
errors. Wrap each deployment with withDeployRetry (2 retries) so
individual failures are retried without rerunning the entire pipeline.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: claude

.dagger/src/index-release-helpers.ts

@@ -129,6 +129,26 @@ type DeployTask = {
   deploy: () => Promise<string>;
 };
 
+async function withDeployRetry(
+  name: string,
+  fn: () => Promise<string>,
+  maxRetries = 2,
+): Promise<string> {
+  for (let attempt = 0; attempt <= maxRetries; attempt++) {
+    try {
+      return await fn();
+    } catch (error) {
+      const msg = error instanceof Error ? error.message : String(error);
+      const isGraphqlError = msg.includes("unknown error while requesting data via graphql");
+      if (!isGraphqlError || attempt === maxRetries) {
+        throw error;
+      }
+      console.log(`⟳ ${name}: graphql error on attempt ${String(attempt + 1)}, retrying...`);
+    }
+  }
+  throw new Error("unreachable");
+}
+
 export async function runAppDeployments(
   options: ReleasePhaseOptions,
 ): Promise<{ outputs: string[]; errors: string[]; appVersions: Record<string, string> }> {
@@ -204,10 +224,10 @@ export async function runAppDeployments(
     }
   }
 
-  // Run all deployment tasks in parallel
+  // Run all deployment tasks in parallel, with per-task GraphQL error retry
   const results = await runNamedParallel(tasks.map(t => ({
     name: t.name,
-    operation: t.deploy,
+    operation: () => withDeployRetry(t.name, t.deploy),
   })));
 
   // Collect results and build appVersions from SUCCESSFUL deployments only