fix(workflow-tests): never read a mock error past the cap as an empty output

An error longer than the 10,000-character cap failed the error branch and
fell through to the output branch, which still read it as an empty output.
The output branch is now strict, so such a reply is refused instead.

Author: waleedlatif1

apps/sim/lib/workflow-tests/protocol.test.ts

@@ -25,4 +25,9 @@ describe('testReplyArgsSchema', () => {
       output: null,
     })
   })
+
+  it('never reads an error past the cap as an empty output', () => {
+    const reply = { runId: 'r', callId: 'c', error: 'x'.repeat(10_001) }
+    expect(testReplyArgsSchema.safeParse(reply).success).toBe(false)
+  })
 })

apps/sim/lib/workflow-tests/protocol.ts

@@ -47,7 +47,10 @@ export const testStartArgsSchema = z.object({
  */
 export const testReplyArgsSchema = z.union([
   z.object({ runId: z.string().min(1), callId: z.string().min(1), error: z.string().max(10_000) }),
-  z.object({ runId: z.string().min(1), callId: z.string().min(1), output: z.unknown() }),
+  // Strict, so a reply carrying an `error` (one past the cap included) is never read as an output.
+  z
+    .object({ runId: z.string().min(1), callId: z.string().min(1), output: z.unknown() })
+    .strict(),
 ])
 
 export const testNextArgsSchema = z.object({ runId: z.string().min(1) })