vision: real camera frames to the model (image_url), auto-detected for multimodal ollama models; --vision/--no-vision flags; verified gemma4:12b sees frames and completed the beacon mission

testbed/chat.py

@@ -62,9 +62,12 @@ async def chat_loop(client: AICCClient, manifest, args: argparse.Namespace) -> i
         model=model,
         system_prompt=CHAT_MISSION,
         log=lambda role, msg: print(f"  [{role}] {msg}"),
+        multimodal=args.vision,
     )
     auto_steps = args.auto_steps
     print(f"[chat] model: {model}  (endpoint {args.base_url})")
+    if controller.multimodal:
+        print("[chat] vision: ON — the model sees the actual camera frames")
     print("[chat] type your commands; /help for the command list; /exit to quit\n")
 
     async def cmd_state() -> None:
@@ -413,6 +416,12 @@ def main() -> int:
         default=3,
         help="how many consecutive nudges a mission may use before giving up (default 3)",
     )
+    parser.add_argument(
+        "--vision",
+        action=argparse.BooleanOptionalAction,
+        default=None,
+        help="pass real camera frames to the model as images (auto-detected for local ollama)",
+    )
     args = parser.parse_args()
     try:
         return asyncio.run(run(args))