wcgw 1.5.1__tar.gz → 1.5.2__tar.gz

{wcgw-1.5.1 → wcgw-1.5.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: wcgw
-Version: 1.5.1
+Version: 1.5.2
 Summary: What could go wrong giving full shell access to chatgpt?
 Project-URL: Homepage, https://github.com/rusiaaman/wcgw
 Author-email: Aman Rusia <gapypi@arcfu.com>
@@ -29,7 +29,9 @@ Description-Content-Type: text/markdown
 
 # Shell and Coding agent on Chatgpt and Claude desktop apps
 
-A custom gpt on chatgpt web/desktop apps to interact with your local shell, edit files, run code, etc.
+- An MCP server on claude desktop for autonomous shell, coding and desktop control agent.
+- A custom gpt on chatgpt web/desktop apps to interact with your local shell, edit files, run code, etc.
+
 
 [![Tests](https://github.com/rusiaaman/wcgw/actions/workflows/python-tests.yml/badge.svg?branch=main)](https://github.com/rusiaaman/wcgw/actions/workflows/python-tests.yml)
 [![Build](https://github.com/rusiaaman/wcgw/actions/workflows/python-publish.yml/badge.svg)](https://github.com/rusiaaman/wcgw/actions/workflows/python-publish.yml)
@@ -40,6 +42,7 @@ A custom gpt on chatgpt web/desktop apps to interact with your local shell, edit
 ### 🚀 Highlights
 
 - ⚡ **Full Shell Access**: No restrictions, complete control.
+- ⚡ **Desktop control on Claude**: Screen capture, mouse control, keyboard control on claude desktop (on mac with docker linux)
 - ⚡ **Create, Execute, Iterate**: Ask the gpt to keep running compiler checks till all errors are fixed, or ask it to keep checking for the status of a long running command till it's done.
 - ⚡ **Interactive Command Handling**: Supports interactive commands using arrow keys, interrupt, and ansi escape sequences.
 - ⚡ **REPL support**: [beta] Supports python/node and other REPL execution.

{wcgw-1.5.1 → wcgw-1.5.2}/README.md

@@ -1,6 +1,8 @@
 # Shell and Coding agent on Chatgpt and Claude desktop apps
 
-A custom gpt on chatgpt web/desktop apps to interact with your local shell, edit files, run code, etc.
+- An MCP server on claude desktop for autonomous shell, coding and desktop control agent.
+- A custom gpt on chatgpt web/desktop apps to interact with your local shell, edit files, run code, etc.
+
 
 [![Tests](https://github.com/rusiaaman/wcgw/actions/workflows/python-tests.yml/badge.svg?branch=main)](https://github.com/rusiaaman/wcgw/actions/workflows/python-tests.yml)
 [![Build](https://github.com/rusiaaman/wcgw/actions/workflows/python-publish.yml/badge.svg)](https://github.com/rusiaaman/wcgw/actions/workflows/python-publish.yml)
@@ -11,6 +13,7 @@ A custom gpt on chatgpt web/desktop apps to interact with your local shell, edit
 ### 🚀 Highlights
 
 - ⚡ **Full Shell Access**: No restrictions, complete control.
+- ⚡ **Desktop control on Claude**: Screen capture, mouse control, keyboard control on claude desktop (on mac with docker linux)
 - ⚡ **Create, Execute, Iterate**: Ask the gpt to keep running compiler checks till all errors are fixed, or ask it to keep checking for the status of a long running command till it's done.
 - ⚡ **Interactive Command Handling**: Supports interactive commands using arrow keys, interrupt, and ansi escape sequences.
 - ⚡ **REPL support**: [beta] Supports python/node and other REPL execution.

{wcgw-1.5.1 → wcgw-1.5.2}/pyproject.toml

@@ -1,7 +1,7 @@
 [project]
 authors = [{ name = "Aman Rusia", email = "gapypi@arcfu.com" }]
 name = "wcgw"
-version = "1.5.1"
+version = "1.5.2"
 description = "What could go wrong giving full shell access to chatgpt?"
 readme = "README.md"
 requires-python = ">=3.11, <3.13"

{wcgw-1.5.1 → wcgw-1.5.2}/src/wcgw/client/anthropic_client.py

@@ -223,9 +223,10 @@ def loop(
             input_schema=GetScreenInfo.model_json_schema(),
             name="GetScreenInfo",
             description="""
-- Get display information of an OS running on docker using image "ghcr.io/anthropics/anthropic-quickstarts:computer-use-demo-latest"
-- If user hasn't provided docker image id, check using `docker ps` and provide the id.
 - Important: call this first in the conversation before ScreenShot, Mouse, and Keyboard tools.
+- Get display information of a linux os running on docker using image "ghcr.io/anthropics/anthropic-quickstarts:computer-use-demo-latest"
+- If user hasn't provided docker image id, check using `docker ps` and provide the id.
+- If the docker is not running, run using `docker run -d -p 6080:6080 ghcr.io/anthropics/anthropic-quickstarts:computer-use-demo-latest`
 - Connects shell to the docker environment.
 - Note: once this is called, the shell enters the docker environment. All bash commands will run over there.
 """,
@@ -234,26 +235,22 @@ def loop(
             input_schema=ScreenShot.model_json_schema(),
             name="ScreenShot",
             description="""
-- Capture screenshot of an OS running on docker using image "ghcr.io/anthropics/anthropic-quickstarts:computer-use-demo-latest"
-- If user hasn't provided docker image id, check using `docker ps` and provide the id.
-- Capture ScreenShot of the current screen for automation.
+- Capture screenshot of the linux os on docker.
 """,
         ),
         ToolParam(
             input_schema=Mouse.model_json_schema(),
             name="Mouse",
             description="""
-- Interact with docker container running image "ghcr.io/anthropics/anthropic-quickstarts:computer-use-demo-latest"
-- If user hasn't provided docker image id, check using `docker ps` and provide the id.
-- Interact with the screen using mouse
+- Interact with the linux os on docker using mouse.
+- Uses xdotool
 """,
         ),
         ToolParam(
             input_schema=Keyboard.model_json_schema(),
             name="Keyboard",
             description="""
-- Interact with docker container running image "ghcr.io/anthropics/anthropic-quickstarts:computer-use-demo-latest"
-- If user hasn't provided docker image id, check using `docker ps` and provide the id.
+- Interact with the linux os on docker using keyboard.
 - Emulate keyboard input to the screen
 - Uses xdootool to send keyboard input, keys like Return, BackSpace, Escape, Page_Up, etc. can be used.
 - Do not use it to interact with Bash tool.

{wcgw-1.5.1 → wcgw-1.5.2}/src/wcgw/client/computer_use.py

@@ -186,6 +186,7 @@ class ComputerTool:
         docker_image_id: Optional[str] = None,
         text: str | None = None,
         coordinate: tuple[int, int] | None = None,
+        do_left_click_on_move: bool | None = None,
         **kwargs: Any,
     ) -> ToolResult:
         if action == "get_screen_info":
@@ -217,7 +218,12 @@ class ComputerTool:
             )
 
             if action == "mouse_move":
-                return self.shell(f"{self.xdotool} mousemove  {x} {y}")
+                if not do_left_click_on_move:
+                    return self.shell(f"{self.xdotool} mousemove  {x} {y}")
+                else:
+                    return self.shell(
+                        f"{self.xdotool} mousemove  {x} {y} click 1",
+                    )
             elif action == "left_click_drag":
                 return self.shell(
                     f"{self.xdotool} mousedown 1 mousemove  {x} {y} mouseup 1",
@@ -401,6 +407,7 @@ def run_computer_tool(
             result = Computer(
                 action="mouse_move",
                 coordinate=(action.action.x, action.action.y),
+                do_left_click_on_move=action.action.do_left_click_on_move,
             )
         elif isinstance(action.action, LeftClickDrag):
             result = Computer(

{wcgw-1.5.1 → wcgw-1.5.2}/src/wcgw/client/mcp_server/Readme.md

@@ -1,5 +1,9 @@
 # Claude desktop support
 
+`wcgw` enables Claude desktop app on Mac to access shell and file system in order to automate tasks, run code, etc.
+
+It also has a computer use feature to connect to linux running on docker. Claude can fully control it including mouse and keyboard.
+
 ## Setup
 
 Update `claude_desktop_config.json` (~/Library/Application Support/Claude/claude_desktop_config.json)
@@ -32,21 +36,10 @@ Computer use is enabled by default. Claude will be able to connect to any docker
 First run a sample docker image with desktop and optionally VNC connection:
 
 ```sh
-docker run \
-    --entrypoint "" \
-    -p 6080:6080 \
-    -e WIDTH=1024 \
-    -e HEIGHT=768 \
-    -d \
-    ghcr.io/anthropics/anthropic-quickstarts:computer-use-demo-latest \
-    bash -c "\
-        ./start_all.sh && \
-        ./novnc_startup.sh && \
-        python http_server.py > /tmp/server_logs.txt 2>&1 & \
-        tail -f /dev/null"
+docker run -p 6080:6080 ghcr.io/anthropics/anthropic-quickstarts:computer-use-demo-latest
 ```
 
-Connect to `http://localhost:6080/vnc.html` for desktop view (VNC) of the system running in the docker. Then ask claude to control the docker os.
+Connect to `http://localhost:6080/vnc.html` for desktop view (VNC) of the system running in the docker. Then ask claude desktop app to control the docker os.
 
 ## Usage
 

{wcgw-1.5.1 → wcgw-1.5.2}/src/wcgw/client/mcp_server/server.py

@@ -151,9 +151,10 @@ async def handle_list_tools() -> list[types.Tool]:
             inputSchema=GetScreenInfo.model_json_schema(),
             name="GetScreenInfo",
             description="""
-- Get display information of an OS running on docker using image "ghcr.io/anthropics/anthropic-quickstarts:computer-use-demo-latest"
-- If user hasn't provided docker image id, check using `docker ps` and provide the id.
 - Important: call this first in the conversation before ScreenShot, Mouse, and Keyboard tools.
+- Get display information of a linux os running on docker using image "ghcr.io/anthropics/anthropic-quickstarts:computer-use-demo-latest"
+- If user hasn't provided docker image id, check using `docker ps` and provide the id.
+- If the docker is not running, run using `docker run -d -p 6080:6080 ghcr.io/anthropics/anthropic-quickstarts:computer-use-demo-latest`
 - Connects shell to the docker environment.
 - Note: once this is called, the shell enters the docker environment. All bash commands will run over there.
 """,
@@ -162,29 +163,26 @@ async def handle_list_tools() -> list[types.Tool]:
             inputSchema=ScreenShot.model_json_schema(),
             name="ScreenShot",
             description="""
-- Capture screenshot of an OS running on docker using image "ghcr.io/anthropics/anthropic-quickstarts:computer-use-demo-latest"
-- If user hasn't provided docker image id, check using `docker ps` and provide the id.
-- Capture ScreenShot of the current screen for automation.
+- Capture screenshot of the linux os on docker.
 """,
         ),
         ToolParam(
             inputSchema=Mouse.model_json_schema(),
             name="Mouse",
             description="""
-- Interact with docker container running image "ghcr.io/anthropics/anthropic-quickstarts:computer-use-demo-latest"
-- If user hasn't provided docker image id, check using `docker ps` and provide the id.
-- Interact with the screen using mouse
+- Interact with the linux os on docker using mouse.
+- Uses xdotool
 """,
         ),
         ToolParam(
             inputSchema=Keyboard.model_json_schema(),
             name="Keyboard",
             description="""
-- Interact with docker container running image "ghcr.io/anthropics/anthropic-quickstarts:computer-use-demo-latest"
-- If user hasn't provided docker image id, check using `docker ps` and provide the id.
+- Interact with the linux os on docker using keyboard.
 - Emulate keyboard input to the screen
 - Uses xdootool to send keyboard input, keys like Return, BackSpace, Escape, Page_Up, etc. can be used.
 - Do not use it to interact with Bash tool.
+- Make sure you've selected a text area or an editable element before sending text.
 """,
         ),
     ]

{wcgw-1.5.1 → wcgw-1.5.2}/src/wcgw/client/tools.py

@@ -170,6 +170,7 @@ def initial_info() -> str:
 System: {uname_sysname}
 Machine: {uname_machine}
 Current working directory: {CWD}
+wcgw version: {importlib.metadata.version("wcgw")}
 """
 
 

{wcgw-1.5.1 → wcgw-1.5.2}/src/wcgw/types_.py

@@ -60,12 +60,12 @@ class GetScreenInfo(BaseModel):
 
 class ScreenShot(BaseModel):
     type: Literal["ScreenShot"]
-    docker_image_id: str
 
 
 class MouseMove(BaseModel):
     x: int
     y: int
+    do_left_click_on_move: bool
     type: Literal["MouseMove"]