Skip to content

Agent Zero AI framework

browser use upgrade

frdel43613d8f80c0
diff --git a/models.py b/models.py
index 879ee3086..f1b8838e9 100644
--- a/models.py
+++ b/models.py
@@ -19,7 +19,7 @@ import litellm
import openai
from python.helpers import dotenv
-from python.helpers import settings
+from python.helpers import settings, dirty_json
from python.helpers.dotenv import load_dotenv
from python.helpers.providers import get_provider_config
from python.helpers.rate_limiter import RateLimiter
@@ -545,7 +545,27 @@ class LiteLLMChatWrapper(SimpleChatModel):
await asyncio.sleep(retry_delay_s)
-class BrowserCompatibleChatWrapper(LiteLLMChatWrapper):
+class AsyncAIChatReplacement:
+ class _Completions:
+ def __init__(self, wrapper):
+ self._wrapper = wrapper
+
+ async def create(self, *args, **kwargs):
+ # call the async _acall method on the wrapper
+ return await self._wrapper._acall(*args, **kwargs)
+
+ class _Chat:
+ def __init__(self, wrapper):
+ self.completions = AsyncAIChatReplacement._Completions(wrapper)
+
+ def __init__(self, wrapper, *args, **kwargs):
+ self._wrapper = wrapper
+ self.chat = AsyncAIChatReplacement._Chat(wrapper)
+
+
+from browser_use.llm import ChatOllama, ChatOpenRouter, ChatGoogle, ChatAnthropic, ChatGroq, ChatOpenAI
+
+class BrowserCompatibleChatWrapper(ChatOpenRouter):
"""
A wrapper for browser agent that can filter/sanitize messages
before sending them to the LLM.
@@ -553,32 +573,61 @@ class BrowserCompatibleChatWrapper(LiteLLMChatWrapper):
def __init__(self, *args, **kwargs):
turn_off_logging()
- super().__init__(*args, **kwargs)
+ # Create the underlying LiteLLM wrapper
+ self._wrapper = LiteLLMChatWrapper(*args, **kwargs)
# Browser-use may expect a 'model' attribute
- self.model = self.model_name
+ self.model = self._wrapper.model_name
+ self.kwargs = self._wrapper.kwargs
- def _call(
- self,
- messages: List[BaseMessage],
- stop: Optional[List[str]] = None,
- run_manager: Optional[CallbackManagerForLLMRun] = None,
- **kwargs: Any,
- ) -> str:
- turn_off_logging()
- result = super()._call(messages, stop, run_manager, **kwargs)
- return result
+ @property
+ def model_name(self) -> str:
+ return self._wrapper.model_name
- async def _astream(
+ @property
+ def provider(self) -> str:
+ return self._wrapper.provider
+
+ def get_client(self, *args, **kwargs): # type: ignore
+ return AsyncAIChatReplacement(self, *args, **kwargs)
+
+ async def _acall(
self,
messages: List[BaseMessage],
stop: Optional[List[str]] = None,
- run_manager: Optional[AsyncCallbackManagerForLLMRun] = None,
+ run_manager: Optional[CallbackManagerForLLMRun] = None,
**kwargs: Any,
- ) -> AsyncIterator[ChatGenerationChunk]:
- turn_off_logging()
- async for chunk in super()._astream(messages, stop, run_manager, **kwargs):
- yield chunk
+ ):
+ # Apply rate limiting if configured
+ apply_rate_limiter_sync(self._wrapper.a0_model_conf, str(messages))
+ # Call the model
+ try:
+ model = kwargs.pop("model", None)
+ kwrgs = {**self._wrapper.kwargs, **kwargs}
+
+ # hack from browser-use to fix json schema for gemini
+ if "response_format" in kwrgs and "json_schema" in kwrgs["response_format"] and model.startswith("gemini/"):
+ kwrgs["response_format"]["json_schema"] = ChatGoogle("")._fix_gemini_schema(self._wrapper.kwargs)
+
+ resp = await acompletion(
+ model=self._wrapper.model_name,
+ messages=messages,
+ stop=stop,
+ **kwrgs,
+ )
+ except Exception as e:
+ raise e
+
+ # another hack for browser-use post process invalid jsons
+ try:
+ if "response_format" in kwrgs and "json_schema" in kwrgs["response_format"] or "json_object" in kwrgs["response_format"]:
+ if resp.choices[0].message.content is not None and not resp.choices[0].message.content.startswith("{"): # type: ignore
+ js = dirty_json.parse(resp.choices[0].message.content) # type: ignore
+ resp.choices[0].message.content = dirty_json.stringify(js) # type: ignore
+ except Exception as e:
+ pass
+
+ return resp
class LiteLLMEmbeddingWrapper(Embeddings):
model_name: str
diff --git a/python/tools/browser_agent.py b/python/tools/browser_agent.py
index 5081b668e..58d88a44e 100644
--- a/python/tools/browser_agent.py
+++ b/python/tools/browser_agent.py
@@ -148,6 +148,7 @@ class State:
),
controller=controller,
enable_memory=False, # Disable memory to avoid state conflicts
+ llm_timeout=3000, # TODO rem
sensitive_data=cast(dict[str, str | dict[str, str]] | None, secrets_dict or {}), # Pass secrets
)
except Exception as e:
@@ -387,7 +388,7 @@ class BrowserAgent(Tool):
def get_use_agent_log(use_agent: browser_use.Agent | None):
result = ["🚦 Starting task"]
if use_agent:
- action_results = use_agent.state.history.action_results() or []
+ action_results = use_agent.history.action_results() or []
short_log = []
for item in action_results:
# final results
diff --git a/requirements.txt b/requirements.txt
index 0e964bda6..17856f8b7 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,6 +1,6 @@
a2wsgi==1.10.8
ansio==0.0.1
-browser-use==0.2.5
+browser-use==0.5.11
docker==7.1.0
duckduckgo-search==6.1.12
faiss-cpu==1.11.0
@@ -19,7 +19,7 @@ langchain-unstructured[all-docs]==0.1.6
openai-whisper==20240930
lxml_html_clean==0.3.1
markdown==3.7
-mcp==1.12.4
+mcp==1.13.1
newspaper3k==0.2.8
paramiko==3.5.0
playwright==1.52.0
diff --git a/run_ui.py b/run_ui.py
index 4066c3328..adbf75085 100644
--- a/run_ui.py
+++ b/run_ui.py
@@ -1,3 +1,4 @@
+import asyncio
from datetime import timedelta
import os
import secrets
@@ -8,6 +9,7 @@ import struct
from functools import wraps
import threading
from flask import Flask, request, Response, session, redirect, url_for, render_template_string
+from werkzeug.wrappers.response import Response as BaseResponse
import initialize
from python.helpers import files, git, mcp_server, fasta2a_server
from python.helpers.files import get_abs_path
@@ -217,7 +219,7 @@ def run():
name = handler.__module__.split(".")[-1]
instance = handler(app, lock)
- async def handler_wrap():
+ async def handler_wrap() -> BaseResponse:
return await instance.handle_request(request=request)
if handler.requires_loopback():