add: tests.json for RunPod deployment validation

Define explicit test cases for RunPod's testing phase:
- Basic inference test to verify worker startup
- Structured outputs test to validate new API functionality

This should resolve the indefinite 'testing' phase hang by providing
concrete pass/fail criteria instead of relying only on health checks.

Author: endolith

tests.json

@@ -0,0 +1,52 @@
+{
+  "tests": [
+    {
+      "name": "basic_inference",
+      "input": {
+        "prompt": "Say hello in one word.",
+        "sampling_params": {
+          "max_tokens": 10,
+          "temperature": 0.1
+        }
+      },
+      "timeout": 120000
+    },
+    {
+      "name": "structured_outputs_json_schema",
+      "input": {
+        "messages": [
+          {
+            "role": "system",
+            "content": "You are a helpful assistant."
+          },
+          {
+            "role": "user",
+            "content": "Analyze: The Eiffel Tower was completed in 1889 for the World's Fair."
+          }
+        ],
+        "sampling_params": {
+          "max_tokens": 200,
+          "temperature": 0.1,
+          "extra_body": {
+            "structured_outputs": {
+              "json": {
+                "type": "object",
+                "properties": {
+                  "reasoning": {"type": "string"},
+                  "question": {"type": "string"}
+                },
+                "required": ["reasoning", "question"]
+              }
+            }
+          }
+        }
+      },
+      "timeout": 120000
+    }
+  ],
+  "config": {
+    "gpuTypeId": "NVIDIA RTX A4000",
+    "gpuCount": 1,
+    "env": []
+  }
+}