minitap-mobile-use 2.1.0__py3-none-any.whl → 2.2.0__py3-none-any.whl

minitap/mobile_use/agents/cortex/cortex.md

@@ -38,7 +38,7 @@ Focus on the **current PENDING subgoal and the next subgoals not yet started**.
   2.2. Otherwise, output a **stringified structured set of instructions** that an **Executor agent** can perform on a real mobile device:
 
 - These must be **concrete low-level actions**.
-- The executor has the following available tools: **{{ executor_tools_list }}**.
+- The executor has the following available tools: {{ executor_tools_list }}.
 - Your goal is to achieve subgoals **fast** - so you must put as much actions as possible in your instructions to complete all achievable subgoals (based on your observations) in one go.
 - To open URLs/links directly, use the `open_link` tool - it will automatically handle opening in the appropriate browser. It also handles deep links.
 - When you need to open an app, use the `find_packages` low-level action to try and get its name. Then, simply use the `launch_app` low-level action to launch it.

minitap/mobile_use/agents/cortex/cortex.py

@@ -46,7 +46,7 @@ class CortexNode:
             current_subgoal=get_current_subgoal(state.subgoal_plan),
             agents_thoughts=state.agents_thoughts,
             executor_feedback=executor_feedback,
-            executor_tools_list=format_tools_list(self.ctx, EXECUTOR_WRAPPERS_TOOLS),
+            executor_tools_list=format_tools_list(ctx=self.ctx, wrappers=EXECUTOR_WRAPPERS_TOOLS),
         )
         messages = [
             SystemMessage(content=system_message),

minitap/mobile_use/agents/executor/executor.py

@@ -3,6 +3,8 @@ from pathlib import Path
 from jinja2 import Template
 from langchain_core.messages import HumanMessage, SystemMessage
 from langchain_google_genai import ChatGoogleGenerativeAI
+from langchain_google_vertexai.chat_models import ChatVertexAI
+
 from minitap.mobile_use.constants import EXECUTOR_MESSAGES_KEY
 from minitap.mobile_use.context import MobileUseContext
 from minitap.mobile_use.graph.state import State
@@ -56,7 +58,7 @@ class ExecutorNode:
         }
 
         # ChatGoogleGenerativeAI does not support the "parallel_tool_calls" keyword
-        if not isinstance(llm, ChatGoogleGenerativeAI):
+        if not isinstance(llm, ChatGoogleGenerativeAI | ChatVertexAI):
             llm_bind_tools_kwargs["parallel_tool_calls"] = True
 
         llm = llm.bind_tools(**llm_bind_tools_kwargs)

minitap/mobile_use/agents/planner/planner.md

@@ -13,7 +13,7 @@ You work like an agile tech lead: defining the key milestones without locking in
    - Don't assume the full UI is visible yet. Plan based on how most mobile apps work, and keep flexibility.
    - List of agents thoughts is empty which is expected, since it is the first plan.
    - Avoid too granular UI actions based tasks (e.g. "tap", "swipe", "copy", "paste") unless explicitly required.
-   - The executor has the following available tools: **{{ executor_tools_list }}**.
+   - The executor has the following available tools: {{ executor_tools_list }}.
      When one of these tools offers a direct shortcut (e.g. `openLink` instead of manually launching a browser and typing a URL), prefer it over decomposed manual steps.
 
 2. **Replanning**

minitap/mobile_use/agents/planner/planner.py

@@ -30,7 +30,10 @@ class PlannerNode:
 
         system_message = Template(
             Path(__file__).parent.joinpath("planner.md").read_text(encoding="utf-8")
-        ).render(platform=self.ctx.device.mobile_platform.value)
+        ).render(
+            platform=self.ctx.device.mobile_platform.value,
+            executor_tools_list=format_tools_list(ctx=self.ctx, wrappers=EXECUTOR_WRAPPERS_TOOLS),
+        )
         human_message = Template(
             Path(__file__).parent.joinpath("human.md").read_text(encoding="utf-8")
         ).render(
@@ -38,7 +41,6 @@ class PlannerNode:
             initial_goal=state.initial_goal,
             previous_plan="\n".join(str(s) for s in state.subgoal_plan),
             agent_thoughts="\n".join(state.agents_thoughts),
-            executor_tools_list=format_tools_list(self.ctx, EXECUTOR_WRAPPERS_TOOLS),
         )
         messages = [
             SystemMessage(content=system_message),

minitap/mobile_use/config.py

@@ -3,7 +3,9 @@ import os
 from pathlib import Path
 from typing import Annotated, Any, Literal
 
+import google.auth
 from dotenv import load_dotenv
+from google.auth.exceptions import DefaultCredentialsError
 from pydantic import BaseModel, Field, SecretStr, ValidationError, model_validator
 from pydantic_settings import BaseSettings
 
@@ -88,7 +90,7 @@ def record_events(output_path: Path | None, events: list[str] | BaseModel | Any)
 
 ### LLM Configuration
 
-LLMProvider = Literal["openai", "google", "openrouter", "xai"]
+LLMProvider = Literal["openai", "google", "openrouter", "xai", "vertexai"]
 LLMUtilsNode = Literal["outputter", "hopper"]
 AgentNode = Literal["planner", "orchestrator", "cortex", "executor"]
 AgentNodeWithFallback = Literal["cortex"]
@@ -98,6 +100,17 @@ DEFAULT_LLM_CONFIG_FILENAME = "llm-config.defaults.jsonc"
 OVERRIDE_LLM_CONFIG_FILENAME = "llm-config.override.jsonc"
 
 
+def validate_vertex_ai_credentials():
+    try:
+        _, project = google.auth.default()
+        if not project:
+            raise Exception("VertexAI requires a Google Cloud project to be set.")
+    except DefaultCredentialsError as e:
+        raise Exception(
+            f"VertexAI requires valid Google Application Default Credentials (ADC): {e}"
+        )
+
+
 class LLM(BaseModel):
     provider: LLMProvider
     model: str
@@ -110,6 +123,8 @@ class LLM(BaseModel):
             case "google":
                 if not settings.GOOGLE_API_KEY:
                     raise Exception(f"{name} requires GOOGLE_API_KEY in .env")
+            case "vertexai":
+                validate_vertex_ai_credentials()
             case "openrouter":
                 if not settings.OPEN_ROUTER_API_KEY:
                     raise Exception(f"{name} requires OPEN_ROUTER_API_KEY in .env")

minitap/mobile_use/controllers/mobile_command_controller.py

@@ -311,9 +311,9 @@ def press_key(ctx: MobileUseContext, key: Key, dry_run: bool = False):
 
 
 class WaitTimeout(Enum):
-    SHORT = 500
-    MEDIUM = 1000
-    LONG = 5000
+    SHORT = "500"
+    MEDIUM = "1000"
+    LONG = "5000"
 
 
 def wait_for_animation_to_end(
@@ -327,7 +327,7 @@ def wait_for_animation_to_end(
 def run_flow_with_wait_for_animation_to_end(
     ctx: MobileUseContext, base_flow: list, dry_run: bool = False
 ):
-    base_flow.append({"waitForAnimationToEnd": {"timeout": WaitTimeout.MEDIUM.value}})
+    base_flow.append({"waitForAnimationToEnd": {"timeout": int(WaitTimeout.MEDIUM.value)}})
     return run_flow(ctx, base_flow, dry_run=dry_run)
 
 

minitap/mobile_use/main.py

@@ -35,9 +35,9 @@ async def run_automation(
     if settings.ADB_HOST:
         config.with_adb_server(host=settings.ADB_HOST, port=settings.ADB_PORT)
     if settings.DEVICE_HARDWARE_BRIDGE_BASE_URL:
-        config.with_hw_bridge_base_url(url=settings.DEVICE_HARDWARE_BRIDGE_BASE_URL)
+        config.with_hw_bridge(url=settings.DEVICE_HARDWARE_BRIDGE_BASE_URL)
     if settings.DEVICE_SCREEN_API_BASE_URL:
-        config.with_screen_api_base_url(url=settings.DEVICE_SCREEN_API_BASE_URL)
+        config.with_screen_api(url=settings.DEVICE_SCREEN_API_BASE_URL)
     if graph_config_callbacks:
         config.with_graph_config_callbacks(graph_config_callbacks)
 

minitap/mobile_use/sdk/agent.py

@@ -278,13 +278,12 @@ class Agent:
                         )
 
                 if stream_mode == "updates":
-                    for key, value in payload.items():  # type: ignore
+                    for _, value in payload.items():  # type: ignore node name, node output
                         if value and "agents_thoughts" in value:
                             new_thoughts = value["agents_thoughts"]
                             last_item = new_thoughts[-1] if new_thoughts else None
                             if last_item:
                                 log_agent_thought(
-                                    prefix=key,
                                     agent_thought=last_item,
                                 )
 

minitap/mobile_use/sdk/builders/agent_config_builder.py

@@ -105,7 +105,7 @@ class AgentConfigBuilder:
         self._task_request_defaults = copy.deepcopy(config)
         return self
 
-    def with_hw_bridge_base_url(self, url: str | ApiBaseUrl) -> "AgentConfigBuilder":
+    def with_hw_bridge(self, url: str | ApiBaseUrl) -> "AgentConfigBuilder":
         """
         Set the base URL for the device HW bridge API.
 
@@ -117,7 +117,7 @@ class AgentConfigBuilder:
         self._servers.hw_bridge_base_url = url
         return self
 
-    def with_screen_api_base_url(self, url: str | ApiBaseUrl) -> "AgentConfigBuilder":
+    def with_screen_api(self, url: str | ApiBaseUrl) -> "AgentConfigBuilder":
         """
         Set the base URL for the device screen API.
 

minitap/mobile_use/sdk/utils.py

@@ -2,11 +2,11 @@ import os
 from pathlib import Path
 
 from pydantic import ValidationError
+
 from minitap.mobile_use.config import LLMConfig, deep_merge_llm_config, get_default_llm_config
 from minitap.mobile_use.utils.file import load_jsonc
 from minitap.mobile_use.utils.logger import get_logger
 
-
 logger = get_logger(__name__)
 
 
@@ -24,5 +24,6 @@ def load_llm_config_override(path: Path) -> LLMConfig:
     try:
         return deep_merge_llm_config(default_config, override_config_dict)
     except ValidationError as e:
-        logger.error(f"Invalid LLM config: {e}. Falling back to default config")
+        logger.error(f"Invalid LLM config: {e}")
+        logger.info("Falling back to default config")
         return default_config

minitap/mobile_use/services/llm.py

@@ -1,10 +1,12 @@
 import logging
-from typing import Literal, TypeVar
 from collections.abc import Awaitable, Callable
-from typing import overload
+from typing import Literal, TypeVar, overload
 
+from langchain_core.language_models.chat_models import BaseChatModel
 from langchain_google_genai import ChatGoogleGenerativeAI
+from langchain_google_vertexai import ChatVertexAI
 from langchain_openai import ChatOpenAI
+
 from minitap.mobile_use.config import (
     AgentNode,
     AgentNodeWithFallback,
@@ -32,6 +34,19 @@ def get_google_llm(
     return client
 
 
+def get_vertex_llm(
+    model_name: str = "gemini-2.5-pro",
+    temperature: float = 0.7,
+) -> ChatVertexAI:
+    client = ChatVertexAI(
+        model_name=model_name,
+        max_tokens=None,
+        temperature=temperature,
+        max_retries=2,
+    )
+    return client
+
+
 def get_openai_llm(
     model_name: str = "o3",
     temperature: float = 1,
@@ -75,7 +90,7 @@ def get_llm(
     *,
     use_fallback: bool = False,
     temperature: float = 1,
-): ...
+) -> BaseChatModel: ...
 
 
 @overload
@@ -84,7 +99,7 @@ def get_llm(
     name: AgentNode,
     *,
     temperature: float = 1,
-): ...
+) -> BaseChatModel: ...
 
 
 @overload
@@ -94,7 +109,7 @@ def get_llm(
     *,
     is_utils: Literal[True],
     temperature: float = 1,
-): ...
+) -> BaseChatModel: ...
 
 
 def get_llm(
@@ -103,7 +118,7 @@ def get_llm(
     is_utils: bool = False,
     use_fallback: bool = False,
     temperature: float = 1,
-):
+) -> BaseChatModel:
     llm = (
         ctx.llm_config.get_utils(name)  # type: ignore
         if is_utils
@@ -118,6 +133,8 @@ def get_llm(
         return get_openai_llm(llm.model, temperature)
     elif llm.provider == "google":
         return get_google_llm(llm.model, temperature)
+    elif llm.provider == "vertexai":
+        return get_vertex_llm(llm.model, temperature)
     elif llm.provider == "openrouter":
         return get_openrouter_llm(llm.model, temperature)
     elif llm.provider == "xai":

minitap/mobile_use/tools/index.py

@@ -19,7 +19,7 @@ from minitap.mobile_use.tools.mobile.tap import tap_wrapper
 from minitap.mobile_use.tools.mobile.wait_for_animation_to_end import (
     wait_for_animation_to_end_wrapper,
 )
-from minitap.mobile_use.tools.tool_wrapper import ToolWrapper
+from minitap.mobile_use.tools.tool_wrapper import CompositeToolWrapper, ToolWrapper
 
 EXECUTOR_WRAPPERS_TOOLS = [
     back_wrapper,
@@ -41,18 +41,24 @@ EXECUTOR_WRAPPERS_TOOLS = [
 ]
 
 
-def get_tools_from_wrappers(ctx: MobileUseContext, wrappers: list[ToolWrapper]) -> list[BaseTool]:
-    """Get the tools from the wrappers."""
-    return [wrapper.tool_fn_getter(ctx) for wrapper in wrappers]
+def get_tools_from_wrappers(
+    ctx: "MobileUseContext",
+    wrappers: list[ToolWrapper],
+) -> list[BaseTool]:
+    tools: list[BaseTool] = []
+    for wrapper in wrappers:
+        if ctx.llm_config.get_agent("executor").provider == "vertexai":
+            # The main swipe tool argument structure is not supported by vertexai, we need to split
+            # this tool into multiple tools
+            if wrapper.tool_fn_getter == swipe_wrapper.tool_fn_getter and isinstance(
+                wrapper, CompositeToolWrapper
+            ):
+                tools.extend(wrapper.composite_tools_fn_getter(ctx))
+                continue
 
-
-def format_tools_list(ctx: MobileUseContext, wrappers: list[ToolWrapper]) -> str:
-    return "\n".join([tool.name for tool in get_tools_from_wrappers(ctx, wrappers)])
+        tools.append(wrapper.tool_fn_getter(ctx))
+    return tools
 
 
-def get_tool_wrapper_from_name(name: str) -> ToolWrapper | None:
-    """Get the tool wrapper from the name."""
-    for wrapper in EXECUTOR_WRAPPERS_TOOLS:
-        if wrapper.tool_fn_getter.__name__ == f"get_{name}_tool":
-            return wrapper
-    return None
+def format_tools_list(ctx: MobileUseContext, wrappers: list[ToolWrapper]) -> str:
+    return ", ".join([tool.name for tool in get_tools_from_wrappers(ctx, wrappers)])

minitap/mobile_use/tools/mobile/swipe.py

@@ -1,18 +1,28 @@
+from typing import Annotated
+
 from langchain_core.messages import ToolMessage
 from langchain_core.tools import tool
-from langchain_core.tools.base import InjectedToolCallId
+from langchain_core.tools.base import BaseTool, InjectedToolCallId
 from langgraph.prebuilt import InjectedState
 from langgraph.types import Command
+from pydantic import Field
+
 from minitap.mobile_use.constants import EXECUTOR_MESSAGES_KEY
 from minitap.mobile_use.context import MobileUseContext
-from minitap.mobile_use.controllers.mobile_command_controller import SwipeRequest
+from minitap.mobile_use.controllers.mobile_command_controller import (
+    CoordinatesSelectorRequest,
+    PercentagesSelectorRequest,
+    SwipeDirection,
+    SwipeRequest,
+    SwipeStartEndCoordinatesRequest,
+    SwipeStartEndPercentagesRequest,
+)
 from minitap.mobile_use.controllers.mobile_command_controller import swipe as swipe_controller
 from minitap.mobile_use.graph.state import State
-from minitap.mobile_use.tools.tool_wrapper import ToolWrapper
-from typing import Annotated
+from minitap.mobile_use.tools.tool_wrapper import CompositeToolWrapper
 
 
-def get_swipe_tool(ctx: MobileUseContext):
+def get_swipe_tool(ctx: MobileUseContext) -> BaseTool:
     @tool
     def swipe(
         tool_call_id: Annotated[str, InjectedToolCallId],
@@ -20,9 +30,7 @@ def get_swipe_tool(ctx: MobileUseContext):
         agent_thought: str,
         swipe_request: SwipeRequest,
     ):
-        """
-        Swipes on the screen.
-        """
+        """Swipes on the screen."""
         output = swipe_controller(ctx=ctx, swipe_request=swipe_request)
         has_failed = output is not None
         tool_message = ToolMessage(
@@ -45,8 +53,98 @@ def get_swipe_tool(ctx: MobileUseContext):
     return swipe
 
 
-swipe_wrapper = ToolWrapper(
+def get_composite_swipe_tools(ctx: MobileUseContext) -> list[BaseTool]:
+    """
+    Returns composite swipe tools for use with Vertex AI LLMs.
+    Each tool handles a specific swipe mode to avoid complex Union type issues.
+    """
+
+    @tool
+    def swipe_coordinates(
+        agent_thought: str,
+        tool_call_id: Annotated[str, InjectedToolCallId],
+        state: Annotated[State, InjectedState],
+        start_x: int = Field(description="Start X coordinate in pixels"),
+        start_y: int = Field(description="Start Y coordinate in pixels"),
+        end_x: int = Field(description="End X coordinate in pixels"),
+        end_y: int = Field(description="End Y coordinate in pixels"),
+        duration: int = Field(description="Duration in ms", ge=1, le=10000, default=400),
+    ):
+        """Swipe using pixel coordinates from start position to end position."""
+        swipe_request = SwipeRequest(
+            swipe_mode=SwipeStartEndCoordinatesRequest(
+                start=CoordinatesSelectorRequest(x=start_x, y=start_y),
+                end=CoordinatesSelectorRequest(x=end_x, y=end_y),
+            ),
+            duration=duration,
+        )
+        return get_swipe_tool(ctx=ctx).invoke(
+            input={
+                "tool_call_id": tool_call_id,
+                "state": state,
+                "agent_thought": agent_thought,
+                "swipe_request": swipe_request,
+            }
+        )
+
+    @tool
+    def swipe_percentages(
+        agent_thought: str,
+        tool_call_id: Annotated[str, InjectedToolCallId],
+        state: Annotated[State, InjectedState],
+        start_x_percent: int = Field(description="Start X percent (0-100)", ge=0, le=100),
+        start_y_percent: int = Field(description="Start Y percent (0-100)", ge=0, le=100),
+        end_x_percent: int = Field(description="End X percent (0-100)", ge=0, le=100),
+        end_y_percent: int = Field(description="End Y percent (0-100)", ge=0, le=100),
+        duration: int = Field(description="Duration in ms", ge=1, le=10000, default=400),
+    ):
+        """Swipe using percentage coordinates from start position to end position."""
+        swipe_request = SwipeRequest(
+            swipe_mode=SwipeStartEndPercentagesRequest(
+                start=PercentagesSelectorRequest(
+                    x_percent=start_x_percent, y_percent=start_y_percent
+                ),
+                end=PercentagesSelectorRequest(x_percent=end_x_percent, y_percent=end_y_percent),
+            ),
+            duration=duration,
+        )
+        return get_swipe_tool(ctx=ctx).invoke(
+            input={
+                "tool_call_id": tool_call_id,
+                "state": state,
+                "agent_thought": agent_thought,
+                "swipe_request": swipe_request,
+            }
+        )
+
+    @tool
+    def swipe_direction(
+        agent_thought: str,
+        tool_call_id: Annotated[str, InjectedToolCallId],
+        state: Annotated[State, InjectedState],
+        direction: SwipeDirection,
+        duration: int = Field(description="Duration in ms", ge=1, le=10000, default=400),
+    ):
+        """Swipe in a specific direction across the screen."""
+        swipe_request = SwipeRequest(
+            swipe_mode=direction,
+            duration=duration,
+        )
+        return get_swipe_tool(ctx=ctx).invoke(
+            input={
+                "tool_call_id": tool_call_id,
+                "state": state,
+                "agent_thought": agent_thought,
+                "swipe_request": swipe_request,
+            }
+        )
+
+    return [swipe_coordinates, swipe_percentages, swipe_direction]
+
+
+swipe_wrapper = CompositeToolWrapper(
     tool_fn_getter=get_swipe_tool,
+    composite_tools_fn_getter=get_composite_swipe_tools,
     on_success_fn=lambda: "Swipe is successful.",
     on_failure_fn=lambda: "Failed to swipe.",
 )

minitap/mobile_use/tools/tool_wrapper.py

@@ -2,6 +2,7 @@ from collections.abc import Callable
 
 from langchain_core.tools import BaseTool
 from pydantic import BaseModel
+
 from minitap.mobile_use.context import MobileUseContext
 
 
@@ -9,3 +10,7 @@ class ToolWrapper(BaseModel):
     tool_fn_getter: Callable[[MobileUseContext], BaseTool]
     on_success_fn: Callable[..., str]
     on_failure_fn: Callable[..., str]
+
+
+class CompositeToolWrapper(ToolWrapper):
+    composite_tools_fn_getter: Callable[[MobileUseContext], list[BaseTool]]

minitap/mobile_use/utils/recorder.py

@@ -45,12 +45,5 @@ def record_interaction(ctx: MobileUseContext, response: BaseMessage):
     return "Screenshot recorded successfully"
 
 
-def log_agent_thought(prefix: str, agent_thought: str):
-    if prefix:
-        prefix = prefix[0].upper() + prefix[1:]
-    else:
-        prefix = "New agent thought"
-    logger.info(
-        f"💭 {Fore.LIGHTMAGENTA_EX + Style.BRIGHT}{prefix}{Style.RESET_ALL}: "
-        f"{Fore.LIGHTMAGENTA_EX}{agent_thought}{Style.RESET_ALL}"
-    )
+def log_agent_thought(agent_thought: str):
+    logger.info(f"💭 {Fore.LIGHTMAGENTA_EX}{agent_thought}{Style.RESET_ALL}")

{minitap_mobile_use-2.1.0.dist-info → minitap_mobile_use-2.2.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: minitap-mobile-use
-Version: 2.1.0
+Version: 2.2.0
 Summary: AI-powered multi-agent system that automates real Android and iOS devices through low-level control using LangGraph.
 Author: Pierre-Louis Favreau, Jean-Pierre Lo, Nicolas Dehandschoewercker
 License: MIT License
@@ -24,11 +24,11 @@ License: MIT License
          LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
          OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
          SOFTWARE.
-Requires-Dist: langgraph==0.5.0
+Requires-Dist: langgraph>=0.6.6
 Requires-Dist: adbutils==2.9.3
-Requires-Dist: langchain-google-genai==2.1.5
-Requires-Dist: langchain==0.3.26
-Requires-Dist: langchain-core==0.3.66
+Requires-Dist: langchain-google-genai>=2.1.10
+Requires-Dist: langchain>=0.3.27
+Requires-Dist: langchain-core>=0.3.75
 Requires-Dist: jinja2==3.1.6
 Requires-Dist: python-dotenv==1.1.1
 Requires-Dist: pydantic-settings==2.10.1
@@ -42,6 +42,7 @@ Requires-Dist: fastapi==0.111.0
 Requires-Dist: uvicorn[standard]==0.30.1
 Requires-Dist: colorama>=0.4.6
 Requires-Dist: psutil>=5.9.0
+Requires-Dist: langchain-google-vertexai>=2.0.28
 Requires-Dist: ruff==0.5.3 ; extra == 'dev'
 Requires-Dist: pytest==8.4.1 ; extra == 'dev'
 Requires-Dist: pytest-cov==5.0.0 ; extra == 'dev'
@@ -69,6 +70,10 @@ Description-Content-Type: text/markdown
     <a href="https://x.com/minitap_ai?t=iRWtI497UhRGLeCKYQekig&s=09"><b>Twitter / X</b></a>
 </p>
 
+[![PyPI version](https://img.shields.io/pypi/v/minitap-mobile-use.svg?color=blue)](https://pypi.org/project/minitap-mobile-use/)
+[![Python Version](https://img.shields.io/badge/python-3.12%2B-blue)](https://www.python.org/downloads/)
+[![License](https://img.shields.io/badge/license-MIT-blue)](https://github.com/minitap-ai/mobile-use/blob/main/LICENSE)
+
 </div>
 
 Mobile-use is a powerful, open-source AI agent that controls your Android or IOS device using natural language. It understands your commands and interacts with the UI to perform tasks, from sending messages to navigating complex apps.
@@ -107,11 +112,26 @@ Ready to automate your mobile experience? Follow these steps to get mobile-use u
 
 2.  **(Optional) Customize LLM Configuration:**
     To use different models or providers, create your own LLM configuration file.
+
     ```bash
     cp llm-config.override.template.jsonc llm-config.override.jsonc
     ```
+
     Then, edit `llm-config.override.jsonc` to fit your needs.
 
+    You can also use local LLMs or any other openai-api compatible providers :
+
+    1. Set `OPENAI_BASE_URL` and `OPENAI_API_KEY` in your `.env`
+    2. In your `llm-config.override.jsonc`, set `openai` as the provider for the agent nodes you want, and choose a model supported by your provider.
+
+    > [!NOTE]  
+    > If you want to use Google Vertex AI, you must either:
+    >
+    > - Have credentials configured for your environment (gcloud, workload identity, etc…)
+    > - Store the path to a service account JSON file as the GOOGLE_APPLICATION_CREDENTIALS environment variable
+    >
+    > More information: - [Credential types](https://cloud.google.com/docs/authentication/application-default-credentials#GAC) - [google.auth API reference](https://googleapis.dev/python/google-auth/latest/reference/google.auth.html#module-google.auth)
+
 ### Quick Launch (Docker)
 
 > [!NOTE]  

{minitap_mobile_use-2.1.0.dist-info → minitap_mobile_use-2.2.0.dist-info}/RECORD

@@ -1,10 +1,10 @@
 minitap/mobile_use/__init__.py,sha256=e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855,0
 minitap/mobile_use/agents/contextor/contextor.py,sha256=d7c98771a9173281c1660a2ee965149d4f70ecf7005cf75fa89eb1e9a9f2b660,1673
-minitap/mobile_use/agents/cortex/cortex.md,sha256=69358852474f65144486fc60196b17bdc5f99d2891fbb4cae26fd2331260f6ba,5598
-minitap/mobile_use/agents/cortex/cortex.py,sha256=232dfe2e2dbbfbc34873875ee32cf5f9d4156f6b9afc9d9d9146addd102be40c,4896
+minitap/mobile_use/agents/cortex/cortex.md,sha256=dcf42ec3081d21d1652b601b00276f6f876806855f4337e0c23d99f20a32b88f,5594
+minitap/mobile_use/agents/cortex/cortex.py,sha256=91d672cb6c6b3eb95ab335ec6031d2479446a8d2af087c65f0bd0300799547bb,4909
 minitap/mobile_use/agents/cortex/types.py,sha256=c33f2277752644d2185d84add03a493adaa530096d046d73366ab9121d99b946,361
 minitap/mobile_use/agents/executor/executor.md,sha256=800003af904d346ab0a6b8a205c0332930161a8c997ede584097a40e8f1ac478,2806
-minitap/mobile_use/agents/executor/executor.py,sha256=be92742d91296650288c12e83f8065589d8eab7aeaa977a47c34ed25879a9bb6,2771
+minitap/mobile_use/agents/executor/executor.py,sha256=e740b76c9d2dfa832afbca60ac533f691963ae068c64dd97a96b2c58c80c925d,2850
 minitap/mobile_use/agents/executor/tool_node.py,sha256=2ad729ede393882460ae3d180ac1c0e1ab1688f40b2017220aad1b059f6485c5,3900
 minitap/mobile_use/agents/executor/utils.py,sha256=1a387d30047d3be05b5e550433866abe4388222b3f95d1360847870155ef8f12,368
 minitap/mobile_use/agents/hopper/hopper.md,sha256=2e9333ece8f6b76401ac2cce98ca06a025faa5dba6bacbbc344793ddf42292d0,362
@@ -17,27 +17,27 @@ minitap/mobile_use/agents/outputter/human.md,sha256=6b9b45c640b163554524b1aec4cd
 minitap/mobile_use/agents/outputter/outputter.py,sha256=0539cd1bfa307c6e24136488a0481128da17c37f20128e63388b5c4aea5aae50,2750
 minitap/mobile_use/agents/outputter/test_outputter.py,sha256=4c52988f8f29159657707a0e11500610f3987cac6390b17edec23f09ddfcc0ff,3334
 minitap/mobile_use/agents/planner/human.md,sha256=cb37be2af568918e60238eaa785837178a3ba8f8112de86850d9a62914c18314,222
-minitap/mobile_use/agents/planner/planner.md,sha256=41c973c8d34a6c24c81b275f725fc447d58805dcea1900f4285c5980502fc756,3337
-minitap/mobile_use/agents/planner/planner.py,sha256=0b8c92857847d390d98efb8d1130762883e46ea4dfc354ea3352e45450913771,2686
+minitap/mobile_use/agents/planner/planner.md,sha256=dc13218df28b38fd2139803c9d013f0ba50899b2ae58545a3b8f3d2fd2aad734,3333
+minitap/mobile_use/agents/planner/planner.py,sha256=cb7fa02cf5ae9e6f95646afc1c41642e5f9e58113dc17d51839f5236a33e67b7,2722
 minitap/mobile_use/agents/planner/types.py,sha256=da551df70ae4ae574670bd8d79c4da1a82be88c24216034d6841a83d85c0c43c,1392
 minitap/mobile_use/agents/planner/utils.py,sha256=11731d0382bf88ef77b38aaf75873486f293343cdd048372c57e626c652b4a22,1839
 minitap/mobile_use/agents/summarizer/summarizer.py,sha256=3e1e92f9259c040487992636786f972dffe0a38d309a8c3758c9ff1aeff2f62c,1070
 minitap/mobile_use/clients/device_hardware_client.py,sha256=9593380a7a3df32f02aa22717678c25e91367df26b1743abde9e57aec5dc2474,857
 minitap/mobile_use/clients/ios_client.py,sha256=332bf47ac01bbd5bf6178a59eea7c7a30fed944f30907caea5388178f312d36b,1452
 minitap/mobile_use/clients/screen_api_client.py,sha256=3615dc65d25c38b4d8dc5512f9adb3bcf69dca7a0298a472a6812f604a275c47,2019
-minitap/mobile_use/config.py,sha256=d5bd2bf71f229d57757bf8375b5caf858b4964178507ecb855942bb197646fdd,9399
+minitap/mobile_use/config.py,sha256=9a2435a69c4453275439555d6af721a0c38bf7ca7ef1916598011c54c198c812,9931
 minitap/mobile_use/constants.py,sha256=3acd9d6ade5bc772e902b3473f3ba12ddd04e7306963ca2bae49d1132d89ba46,95
 minitap/mobile_use/context.py,sha256=fa4e43fb580db86c2ed707b1464d8363ead57332b40ea5688e0731ad57a40558,1747
 minitap/mobile_use/controllers/__init__.py,sha256=e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855,0
-minitap/mobile_use/controllers/mobile_command_controller.py,sha256=1ea22f569affae3d82ea74afd37629898cf3d10b3bdf6cb1a7671d1a09f2c000,12206
+minitap/mobile_use/controllers/mobile_command_controller.py,sha256=eb5f31cc6f2a711cbe3908f2561e841abe2214805475f77eeed3bc45731baddc,12217
 minitap/mobile_use/controllers/platform_specific_commands_controller.py,sha256=8b4fc30108c242da41fd998751dbfd5e6a69e2957a2dbbe5d6fc43d6b55f727e,2705
 minitap/mobile_use/graph/graph.py,sha256=c7b412e725b096eca8f212d704c3faf91d77eea4131f1fae7af1ee39bc57cdae,4269
 minitap/mobile_use/graph/state.py,sha256=cfe67d15833efa18b28a86293bc5c713ff8c777778b0552bb6e25fa7070534aa,3371
-minitap/mobile_use/main.py,sha256=51a826974071997d7ca1d6913e199c7844633e483db0471f3c294688be028834,3743
+minitap/mobile_use/main.py,sha256=7ac4dc592e3ce72bff602d67ba2f25b9b5e45e07a316e548d7c8e73735abf43d,3725
 minitap/mobile_use/sdk/__init__.py,sha256=4e5555c0597242b9523827194a2500b9c6d7e5c04b1ccd2056c9b1f4d42a31cd,318
-minitap/mobile_use/sdk/agent.py,sha256=43b198cfabdffa28add7f1b77046704ee28da74da92b710dc8bb265c8df16472,20898
+minitap/mobile_use/sdk/agent.py,sha256=0d598bc0fab7e4429654e38379d730a021243a0953ad8efcca721b382938be7a,20871
 minitap/mobile_use/sdk/builders/__init__.py,sha256=d6c96d39b80900a114698ef205ab5061a541f33bfa99c456d9345e5adb8ff6ff,424
-minitap/mobile_use/sdk/builders/agent_config_builder.py,sha256=d963ab932fa78117eef3ea4a2503c34b872f7b865c1a2e25b20e7fa5072e466b,7600
+minitap/mobile_use/sdk/builders/agent_config_builder.py,sha256=c63f452350822daab86d29c44a333909b623fc7ff7bcbf74e5f9104b24630bf5,7582
 minitap/mobile_use/sdk/builders/index.py,sha256=64336ac3b3dea4673a48e95b8c5ac4196ecd5d2196380377d102593d0a1dc138,442
 minitap/mobile_use/sdk/builders/task_request_builder.py,sha256=9e6cf7afb68af986d6a81487179bb79d28f63047a068725d92996dbcbe753376,6857
 minitap/mobile_use/sdk/constants.py,sha256=436ba0700c6cf37ac0c9e3995a5f5a0d54ca87af72686eb9667a2c6a96e30f68,292
@@ -49,7 +49,7 @@ minitap/mobile_use/sdk/types/__init__.py,sha256=5dd148d83bf6261ac8ac60c994e2496b
 minitap/mobile_use/sdk/types/agent.py,sha256=390d5c642b3480f4a2203ddd28ec115c785f2576bec81e82e4db3c129399c020,2260
 minitap/mobile_use/sdk/types/exceptions.py,sha256=56ac3f749730740951448b1b0f200be21331dc0800916a87587b21e7850120a5,2288
 minitap/mobile_use/sdk/types/task.py,sha256=74743a398b63af62383528d5906824ae8aaba1e1885c75414b347623d7931f12,5837
-minitap/mobile_use/sdk/utils.py,sha256=493c77e43fcb58535eef43416716a3283488577c127060de5c0317d0b737b01f,945
+minitap/mobile_use/sdk/utils.py,sha256=647f1f4a463c3029c3b0eb3c33f7dd778d5f5fd9d293224f5474595a60e1de6f,967
 minitap/mobile_use/servers/config.py,sha256=8a4a6bce23e2093d047a91e135e2f88627f76ac12177d071f25a3ca739b3afeb,575
 minitap/mobile_use/servers/device_hardware_bridge.py,sha256=80b93fe1bd8ea9100ac198a83f0aea2c40565a11e810acff9785bbd3f3b31f37,7174
 minitap/mobile_use/servers/device_screen_api.py,sha256=63bf866f17cde4ab97631b710080866b8427225d3857b2351ab83db38a9c5107,5064
@@ -57,8 +57,8 @@ minitap/mobile_use/servers/start_servers.py,sha256=1e86dc0fcbdf6e6570ae68c709714
 minitap/mobile_use/servers/stop_servers.py,sha256=9a3dc2eafb3c13e420248b1844694c80112be32f0d336f54ecc1015cb6f27be9,7127
 minitap/mobile_use/servers/utils.py,sha256=db5d26153a169ab141556337db3693adc1bf8522943316656bdeb05dbf95465b,394
 minitap/mobile_use/services/accessibility.py,sha256=42bcbe81b427ee6f6e82bcfe420fc40630db950bda354e3e433c2dda2e159628,3404
-minitap/mobile_use/services/llm.py,sha256=c3e99290431bcdcbf81cbffd08bf3abbd4a3760c20cb0873278c418df0d28b0f,3724
-minitap/mobile_use/tools/index.py,sha256=db61053f24beb88952cc34be37336defd8ebf5e1fec0ecf81e2439a76b29bde1,2449
+minitap/mobile_use/services/llm.py,sha256=3706edcb2132709c5cb931fac86ea31908a209347b341ec7b0f1cfbb27959c66,4260
+minitap/mobile_use/tools/index.py,sha256=b380fa7b756108ca3dd1f81fbe397233f742ddcecaaefa2844c2e67ced7093be,2695
 minitap/mobile_use/tools/mobile/back.py,sha256=cf1053b22c4fbeb1c219578563d6d857425dcdff08af690149c6e52a0e29c195,1792
 minitap/mobile_use/tools/mobile/clear_text.py,sha256=28e9af14cbb62be3a8221e7b1e7645e9bc22f40202d84c6047874809137cd927,9316
 minitap/mobile_use/tools/mobile/copy_text_from.py,sha256=723d7563e19d05060da0071aedfb620ce36bfb90931fcb815ab8d01be7ed9dd5,2716
@@ -71,11 +71,11 @@ minitap/mobile_use/tools/mobile/open_link.py,sha256=ad363f83b254a8e3e5969b4d8b90
 minitap/mobile_use/tools/mobile/paste_text.py,sha256=dc126297f174cdcdfc6f6d148fdb6a96ecb7813926e441ce0631e963b196b448,2101
 minitap/mobile_use/tools/mobile/press_key.py,sha256=a22d416279c33f2e843aaf28e4466a7eeae59aa690312c8866b62a3f84b57d5b,1939
 minitap/mobile_use/tools/mobile/stop_app.py,sha256=8fc1cf5682232d670270a7e909f365a795150643eac5310f042d792b79e7c0c0,2062
-minitap/mobile_use/tools/mobile/swipe.py,sha256=4d0da8c0db0995598a9a74ebfc5e7fdc176abc8e5277fd8d5f588de864947dda,1878
+minitap/mobile_use/tools/mobile/swipe.py,sha256=23c31022bef827afbfe90c7f6a7d195e5bef57eb3972cd32b0d2aa054b4c6d26,5754
 minitap/mobile_use/tools/mobile/take_screenshot.py,sha256=8762be82e0fb55549f1271a8e4c7b25040f906d21ed19f52121a616e70eb9bb0,2271
 minitap/mobile_use/tools/mobile/tap.py,sha256=d7a3de2ddb78b051b22d7886553d75bab13562abfc8562957c15fb16dd484a0a,2297
 minitap/mobile_use/tools/mobile/wait_for_animation_to_end.py,sha256=967d0df11dfb073dde6308761e9795e02c93ee2b0cbe17caf6edbe4a8beea28a,2493
-minitap/mobile_use/tools/tool_wrapper.py,sha256=166d9c4b7950e66407eb397a81d88354779ee3056a8dc7ad1bf160a3aa54580c,334
+minitap/mobile_use/tools/tool_wrapper.py,sha256=f0f27beaae25a1bcfd9b72bf994de84b2e5fba9e242d8ad18a8d1a97cd7619e4,454
 minitap/mobile_use/tools/utils.py,sha256=bf6924e3d9e5338c2e04d88fd0ad9ed03a0c02058b62373025fa4b3a805712ae,2694
 minitap/mobile_use/utils/cli_helpers.py,sha256=1c53b6ea6cd2ba861302b182944c6a3a31dac27e316bca2c65cd6a3ca3256e81,1720
 minitap/mobile_use/utils/cli_selection.py,sha256=62e949bf075e984b5d23b4a9880ff2bccf8f9e0f7ccb48120030a6a82075352b,4788
@@ -85,12 +85,12 @@ minitap/mobile_use/utils/errors.py,sha256=6c5566484cff48ce1eb168c3cbe93d6e536545
 minitap/mobile_use/utils/file.py,sha256=1ca968613452a273b23e4f58460ab39f87255b02cdb6fb8ca04f4e628b346070,315
 minitap/mobile_use/utils/logger.py,sha256=011fe08c39b111997ec685b9f0b378761607e35ac049234b5e86c2b58f29bbe3,5633
 minitap/mobile_use/utils/media.py,sha256=29fa7f009a0f2bd60de2a9eba4a90e6eecc91cc67c2736b753506d48ab2bf5eb,2228
-minitap/mobile_use/utils/recorder.py,sha256=83dcb953b9caf00bf3e4853c3eee447166e1279e7ceefbbeb25d93e932ec698d,1937
+minitap/mobile_use/utils/recorder.py,sha256=f90c511a182553684fc4c51195dec6859c7867b83b66bd56dd13f0cdcab2b3ae,1724
 minitap/mobile_use/utils/requests_utils.py,sha256=5c3a2e2aff7c521cd6a43b74c084e2244e1ff55065a5b722e6e251c6419861fd,1168
 minitap/mobile_use/utils/shell_utils.py,sha256=b35ae7f863379adb86c9ba0f9b3b9d4954118d12aef1ffed0bc260b32d73d857,650
 minitap/mobile_use/utils/time.py,sha256=41bfaabb3751de11443ccb4a3f1f53d5ebacc7744c72e32695fdcc3d23f17d49,160
 minitap/mobile_use/utils/ui_hierarchy.py,sha256=b52acd081ac18169dab94b7187c6ffed4db72d4e4b7f56aeca9ef5b81166c4e0,3630
-minitap_mobile_use-2.1.0.dist-info/WHEEL,sha256=ab6157bc637547491fb4567cd7ddf26b04d63382916ca16c29a5c8e94c9c9ef7,79
-minitap_mobile_use-2.1.0.dist-info/entry_points.txt,sha256=663a29cfd551a4eaa0f27335f0bd7e4a732a4e39c76b68ef5c8dc444d4a285fa,60
-minitap_mobile_use-2.1.0.dist-info/METADATA,sha256=95abc6db29ea323284374508c8faba3fd6cf83db998a2a45b82585af354b8ab2,10574
-minitap_mobile_use-2.1.0.dist-info/RECORD,,
+minitap_mobile_use-2.2.0.dist-info/WHEEL,sha256=ab6157bc637547491fb4567cd7ddf26b04d63382916ca16c29a5c8e94c9c9ef7,79
+minitap_mobile_use-2.2.0.dist-info/entry_points.txt,sha256=663a29cfd551a4eaa0f27335f0bd7e4a732a4e39c76b68ef5c8dc444d4a285fa,60
+minitap_mobile_use-2.2.0.dist-info/METADATA,sha256=b0076a10119155863d990ff779bb901857df8b74c1321d6fef45ac6a16029db8,11825
+minitap_mobile_use-2.2.0.dist-info/RECORD,,