wcgw 2.0.1__tar.gz → 2.0.3__tar.gz

{wcgw-2.0.1 → wcgw-2.0.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: wcgw
-Version: 2.0.1
+Version: 2.0.3
 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>
@@ -27,9 +27,10 @@ Requires-Dist: uvicorn>=0.31.0
 Requires-Dist: websockets>=13.1
 Description-Content-Type: text/markdown
 
-# Shell and Coding agent on Claude desktop app
+# Shell and Coding agent for Claude and Chatgpt
 
-- An MCP server on claude desktop for autonomous shell, coding and desktop control agent.
+- Claude - An MCP server on claude desktop for autonomous shell, coding and desktop control agent.
+- Chatgpt - Allows custom gpt to talk to your shell via a relay server. 
 
 [![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)
 [![Mypy strict](https://github.com/rusiaaman/wcgw/actions/workflows/python-types.yml/badge.svg?branch=main)](https://github.com/rusiaaman/wcgw/actions/workflows/python-types.yml)
@@ -37,7 +38,7 @@ Description-Content-Type: text/markdown
 
 ## Updates
 
-- [01 Dec 2024] Deprecated chatgpt app support
+- [01 Dec 2024] Removed author hosted relay server for chatgpt.
 
 - [26 Nov 2024] Introduced claude desktop support through mcp
 
@@ -49,7 +50,7 @@ Description-Content-Type: text/markdown
 - ⚡ **Interactive Command Handling**: Supports interactive commands using arrow keys, interrupt, and ansi escape sequences.
 - ⚡ **REPL support**: [beta] Supports python/node and other REPL execution.
 
-## Setup
+## Claude Setup
 
 Update `claude_desktop_config.json` (~/Library/Application Support/Claude/claude_desktop_config.json)
 
@@ -74,7 +75,7 @@ Update `claude_desktop_config.json` (~/Library/Application Support/Claude/claude
 
 Then restart claude app.
 
-## [Optional] Computer use support using desktop on docker
+### [Optional] Computer use support using desktop on docker
 
 Computer use is disabled by default. Add `--computer-use` to enable it. This will add necessary tools to Claude including ScreenShot, Mouse and Keyboard control.
 
@@ -123,7 +124,12 @@ Then ask claude to execute shell commands, read files, edit files, run your code
 
 If you've run the docker for LLM to access, you can ask it to control the "docker os". If you don't provide the docker container id to it, it'll try to search for available docker using `docker ps` command.
 
-## Example
+
+## Chatgpt Setup
+
+Read here: https://github.com/rusiaaman/wcgw/blob/main/openai.md
+
+## Examples
 
 ### Computer use example
 
@@ -133,6 +139,7 @@ If you've run the docker for LLM to access, you can ask it to control the "docke
 
 ![example](https://github.com/rusiaaman/wcgw/blob/main/static/example.jpg?raw=true)
 
+
 ## [Optional] Local shell access with openai API key or anthropic API key
 
 ### Openai

{wcgw-2.0.1 → wcgw-2.0.3}/README.md

@@ -1,6 +1,7 @@
-# Shell and Coding agent on Claude desktop app
+# Shell and Coding agent for Claude and Chatgpt
 
-- An MCP server on claude desktop for autonomous shell, coding and desktop control agent.
+- Claude - An MCP server on claude desktop for autonomous shell, coding and desktop control agent.
+- Chatgpt - Allows custom gpt to talk to your shell via a relay server. 
 
 [![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)
 [![Mypy strict](https://github.com/rusiaaman/wcgw/actions/workflows/python-types.yml/badge.svg?branch=main)](https://github.com/rusiaaman/wcgw/actions/workflows/python-types.yml)
@@ -8,7 +9,7 @@
 
 ## Updates
 
-- [01 Dec 2024] Deprecated chatgpt app support
+- [01 Dec 2024] Removed author hosted relay server for chatgpt.
 
 - [26 Nov 2024] Introduced claude desktop support through mcp
 
@@ -20,7 +21,7 @@
 - ⚡ **Interactive Command Handling**: Supports interactive commands using arrow keys, interrupt, and ansi escape sequences.
 - ⚡ **REPL support**: [beta] Supports python/node and other REPL execution.
 
-## Setup
+## Claude Setup
 
 Update `claude_desktop_config.json` (~/Library/Application Support/Claude/claude_desktop_config.json)
 
@@ -45,7 +46,7 @@ Update `claude_desktop_config.json` (~/Library/Application Support/Claude/claude
 
 Then restart claude app.
 
-## [Optional] Computer use support using desktop on docker
+### [Optional] Computer use support using desktop on docker
 
 Computer use is disabled by default. Add `--computer-use` to enable it. This will add necessary tools to Claude including ScreenShot, Mouse and Keyboard control.
 
@@ -94,7 +95,12 @@ Then ask claude to execute shell commands, read files, edit files, run your code
 
 If you've run the docker for LLM to access, you can ask it to control the "docker os". If you don't provide the docker container id to it, it'll try to search for available docker using `docker ps` command.
 
-## Example
+
+## Chatgpt Setup
+
+Read here: https://github.com/rusiaaman/wcgw/blob/main/openai.md
+
+## Examples
 
 ### Computer use example
 
@@ -104,6 +110,7 @@ If you've run the docker for LLM to access, you can ask it to control the "docke
 
 ![example](https://github.com/rusiaaman/wcgw/blob/main/static/example.jpg?raw=true)
 
+
 ## [Optional] Local shell access with openai API key or anthropic API key
 
 ### Openai

{wcgw-2.0.1 → wcgw-2.0.3}/pyproject.toml

@@ -1,7 +1,7 @@
 [project]
 authors = [{ name = "Aman Rusia", email = "gapypi@arcfu.com" }]
 name = "wcgw"
-version = "2.0.1"
+version = "2.0.3"
 description = "What could go wrong giving full shell access to chatgpt?"
 readme = "README.md"
 requires-python = ">=3.11, <3.13"

{wcgw-2.0.1 → wcgw-2.0.3}/src/wcgw/client/anthropic_client.py

@@ -262,6 +262,7 @@ def loop(
 - Interact with the linux os on docker using mouse.
 - Uses xdotool
 - About left_click_drag: the current mouse position will be used as the starting point, click and drag to the given x, y coordinates. Useful in things like sliders, moving things around, etc.
+- The output of this command has the screenshot after doing this action. Use this to verify if the action was successful.
 """,
             ),
             ToolParam(
@@ -272,6 +273,8 @@ def loop(
 - 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.
+- The output of this command has the screenshot after doing this action. Use this to verify if the action was successful.
 """,
             ),
         ]

{wcgw-2.0.1 → wcgw-2.0.3}/src/wcgw/client/computer_use.py

@@ -245,9 +245,17 @@ class ComputerTool:
                 return self.shell(f"{self.xdotool} key -- {text}")
             elif action == "type":
                 results: list[ToolResult] = []
-                for chunk in chunks(text, TYPING_GROUP_SIZE):
-                    cmd = f"{self.xdotool} type --delay {TYPING_DELAY_MS} -- {shlex.quote(chunk)}"
-                    results.append(self.shell(cmd, take_screenshot=False))
+                all_lines = text.splitlines()
+                for i, line in enumerate(all_lines):
+                    for chunk in chunks(line, TYPING_GROUP_SIZE):
+                        cmd = f"{self.xdotool} type --delay {TYPING_DELAY_MS} -- {shlex.quote(chunk)}"
+                        results.append(self.shell(cmd, take_screenshot=False))
+                    if i < len(all_lines) - 1:
+                        results.append(
+                            self.shell(
+                                f"{self.xdotool} key Return", take_screenshot=False
+                            )
+                        )
                 screenshot_base64 = self.screenshot().base64_image
                 return ToolResult(
                     output="".join(result.output or "" for result in results),
@@ -343,8 +351,9 @@ class ComputerTool:
 
     def shell(self, command: str, take_screenshot: bool = True) -> ToolResult:
         """Run a shell command and return the output, error, and optionally a screenshot."""
+        escaped_command = shlex.quote(command)
         _, stdout, stderr = command_run(
-            f"docker exec {self.docker_image_id} sh -c '{command}'"
+            f"docker exec {self.docker_image_id} bash -c {escaped_command}",
         )
         base64_image = None
 

{wcgw-2.0.1 → wcgw-2.0.3}/src/wcgw/client/mcp_server/server.py

@@ -178,6 +178,7 @@ async def handle_list_tools() -> list[types.Tool]:
 - Interact with the linux os on docker using mouse.
 - Uses xdotool
 - About left_click_drag: the current mouse position will be used as the starting point, click and drag to the given x, y coordinates. Useful in things like sliders, moving things around, etc.
+- The output of this command has the screenshot after doing this action. Use this to verify if the action was successful.
 """,
             ),
             ToolParam(
@@ -189,6 +190,7 @@ async def handle_list_tools() -> list[types.Tool]:
 - 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.
+- The output of this command has the screenshot after doing this action. Use this to verify if the action was successful.
 """,
             ),
         ]

{wcgw-2.0.1 → wcgw-2.0.3}/uv.lock

@@ -860,7 +860,7 @@ wheels = [
 
 [[package]]
 name = "wcgw"
-version = "1.5.4"
+version = "2.0.2"
 source = { editable = "." }
 dependencies = [
     { name = "anthropic" },