opencode: add qwen3.8-flash-next to the work profile

Served by a rented MI325X through mainframe's LiteLLM; only answers while
a golden_image session is up.

common/opencode/work.json

@@ -48,6 +48,19 @@
             "cache_write": 0.0
           }
         },
+        "qwen3.8-flash-next": {
+          "name": "qwen3.8-flash-next (rented MI325X; run gi up on mainframe first)",
+          "limit": {
+            "context": 262144,
+            "output": 32768
+          },
+          "cost": {
+            "input": 0.0,
+            "output": 0.0,
+            "cache_read": 0.0,
+            "cache_write": 0.0
+          }
+        },
         "local/librarian": {
           "name": "Librarian - local gpt-oss",
           "limit": {