Open-source Python library for AI browser agents, plus Browser Use Cloud: stealth browsers and a hosted agent API.
Python ai-agentsai-toolsbrowser-agentbrowser-automationbrowser-usecdpcloud-browsercomputer-useheadless-browserllmmcpplaywrightpythonremote-browserstealth-browserweb-automationweb-scraping
fix cubic comment
kalil0321b0c123abed2d
diff --git a/browser_use/llm/vercel/chat.py b/browser_use/llm/vercel/chat.pyindex d7a426677..bba86a026 100644--- a/browser_use/llm/vercel/chat.py+++ b/browser_use/llm/vercel/chat.py@@ -14,7 +14,7 @@ from pydantic import BaseModelfrom browser_use.llm.base import BaseChatModelfrom browser_use.llm.exceptions import ModelProviderError, ModelRateLimitError-from browser_use.llm.messages import BaseMessage+from browser_use.llm.messages import BaseMessage, ContentPartTextParam, SystemMessagefrom browser_use.llm.schema import SchemaOptimizerfrom browser_use.llm.vercel.serializer import VercelMessageSerializerfrom browser_use.llm.views import ChatInvokeCompletion, ChatInvokeUsage@@ -410,12 +410,24 @@ class ChatVercel(BaseChatModel):schema = SchemaOptimizer.create_gemini_optimized_schema(output_format)json_instruction = f'\n\nIMPORTANT: You must respond with ONLY a valid JSON object (no markdown, no code blocks, no explanations) that exactly matches this schema:\n{json.dumps(schema, indent=2)}'+ instruction_added = Falseif modified_messages and modified_messages[0].role == 'system':if isinstance(modified_messages[0].content, str):modified_messages[0].content += json_instruction+ instruction_added = True+ elif isinstance(modified_messages[0].content, list):+ modified_messages[0].content.append(ContentPartTextParam(text=json_instruction))+ instruction_added = Trueelif modified_messages and modified_messages[-1].role == 'user':if isinstance(modified_messages[-1].content, str):modified_messages[-1].content += json_instruction+ instruction_added = True+ elif isinstance(modified_messages[-1].content, list):+ modified_messages[-1].content.append(ContentPartTextParam(text=json_instruction))+ instruction_added = True++ if not instruction_added:+ modified_messages.insert(0, SystemMessage(content=json_instruction))vercel_messages = VercelMessageSerializer.serialize_messages(modified_messages)