Handle stale workspaces: 409 conflict auto-cleanup + verbose logging

When the orchestrator restarts, it loses in-memory queue state. If a
stale workspace still exists, createWorkspace now auto-deletes it and
retries instead of failing with 409. Also adds --verbose to Claude Code
invocation for better task log debugging.

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

coder/cloudflare-worker/main.tf

@@ -579,8 +579,10 @@ LOGEOF
         ARGS="$DEPLOY_ENV"
       fi
 
-      # Run Claude Code in non-interactive mode
+      # Run Claude Code in non-interactive mode with tool access
-      claude --print --dangerously-skip-permissions "/project:$TASK_TYPE $ARGS" 2>&1 | tee ~/task-output.log
+      claude -p --dangerously-skip-permissions --verbose "/project:$TASK_TYPE $ARGS" 2>&1 | tee ~/task-output.log
+      EXIT_CODE=$?
+      echo "Claude exited with code: $EXIT_CODE" | tee -a ~/task-output.log
 
       echo "Task completed. Output saved to ~/task-output.log"
 

src/services/coder.ts

@@ -50,6 +50,18 @@ export class CoderClient {
       },
     );
 
+    // Handle 409 conflict — stale workspace from a previous orchestrator instance
+    if (res.status === 409) {
+      console.log(`[coder] workspace "${name}" already exists — deleting stale workspace and retrying`);
+      const existing = await this.findWorkspaceByName(name);
+      if (existing) {
+        await this.deleteWorkspace(existing.id);
+        // Wait for deletion to propagate
+        await new Promise((resolve) => setTimeout(resolve, 5000));
+      }
+      return this.createWorkspace(task);
+    }
+
     if (!res.ok) {
       const text = await res.text();
       throw new Error(`Coder API error ${res.status}: ${text}`);