fix(vision): stop capping aux vision output with hardcoded max_tokens

The vision tools' call_kwargs hardcode max_tokens caps (2000 for
vision_analyze/browser_vision, 4000 for video analysis), truncating
descriptions of complex images at the cap. The centralized aux client
already omits max_tokens by default (#34845) so providers use their
model max output; these three call sites were the leftovers that
bypassed that policy.

Remove the hardcoded caps entirely — the aux client handles the
mandatory-max_tokens Anthropic wire via _resolve_anthropic_messages_max_tokens
(model output ceiling) and Gemini native omits maxOutputTokens (65K ceiling),
so no wire needs an explicit cap.

tools/browser_tool.py

@@ -4692,7 +4692,6 @@ def browser_vision(question: str, annotate: bool = False, task_id: Optional[str]
                     ],
                 }
             ],
-            "max_tokens": 2000,
             "temperature": vision_temperature,
             "timeout": vision_timeout,
         }

tools/vision_tools.py

@@ -1501,7 +1501,6 @@ async def vision_analyze_tool(
             "task": "vision",
             "messages": messages,
             "temperature": vision_temperature,
-            "max_tokens": 2000,
             "timeout": vision_timeout,
         }
         if model:
@@ -2073,7 +2072,6 @@ async def video_analyze_tool(
             "task": "vision",
             "messages": messages,
             "temperature": vision_temperature,
-            "max_tokens": 4000,
             "timeout": vision_timeout,
         }
         if model: