Toolit

Model Context Protocol (MCP) Server, Typer CLI and Visual Studio Code tasks in one, provides an easy way to configure your own DevTools in a project.

Installation

To get started with Toolit, install the package via pip:

pip install toolit

If you want mcp server support, you can install the optional dependency:

pip install toolit[mcp]

Usage

Add a folder called devtools to your project root. Create python modules, you decide the name, in this folder. Add the tool decorator to functions you want to expose as commands.

# devtools/my_commands.py
from toolit import tool
@tool
def my_command(to_print: str = "Hello, World!") -> None:
    """This is a command that can be run from the CLI."""
    print(to_print)

Toolit will automatically discover these modules and make them available as commands.

Now you can run your command from the command line:

toolit --help  # To see available commands
toolit my-command --to_print "Hello, Toolit!"  # To run your command

You add a command line command with parameter as a devtool, and it will be available as a command in the CLI and in the generated tasks.json file for Visual Studio Code. You do this by:

# devtools/my_cli_commands.py
from toolit import clitool
@clitool
def execute_cli_command(to_print: str = "Hello, World!") -> str:
    """This is a command that can be run from the CLI."""
    return "python -c 'print(\"" + to_print + "\")'"

When this command runs from CLI or a generated VS Code task, Toolit executes the returned string in your shell.

Customizing the DevTools Folder

By default, Toolit looks for a folder named devtools in the project root. You can customize this by creating a toolit.ini or use your pyproject.toml file in your project root with the following content:

[toolit]
tools_folder = "tools"

Create the VS code tasks.json file

You can automatically create a tasks.json file for Visual Studio Code to run your ToolIt commands directly from the editor. This is useful for integrating your development tools into your workflow.

To create the .vscode/tasks.json file, run the following command in your terminal:

toolit create-vscode-tasks-json

NOTE: THIS WILL OVERWRITE YOUR EXISTING .vscode/tasks.json FILE IF IT EXISTS!

List parameter inputs in VS Code tasks

VS Code task inputs do not support native list pickers. For tool parameters typed as list[str], list[int], or list[Enum], Toolit generates a promptString input and expects comma-separated values.

Examples:

• list[str]: alice, bob, charlie
• list[int]: 1, 2, 3
• list[MyEnum]: accepts both enum names and enum values, for example OPTION_A, Option b

Chaining Commands

You can chain multiple using the @sequential_group_of_tools and @parallel_group_of_tools decorators to create more complex workflows. Functions decorated with these decorators should always return a list of callable functions.

from toolit import tool, sequential_group_of_tools, parallel_group_of_tools
from typing import Callable

@tool
def first_command() -> None:
    print("First command executed.")

@tool
def second_command() -> None:
    print("Second command executed.")

@sequential_group_of_tools
def my_sequential_commands() -> list[Callable]:
    return [first_command, second_command]

@parallel_group_of_tools
def my_parallel_commands() -> list[Callable]:
    return [first_command, second_command]

This will create a group of commands in the tasks.json file that can be executed sequentially or in parallel.

Creating Plugins

Toolit supports a plugin system that allows you to create and share your own tools as separate packages. This makes it easy to reuse tools across different projects, without needing to copy and update tools across multiple codebases.

To create a plugin, follow these steps:

1. Create a new Python package for your plugin. You can use tools like setuptools, poetry or uv to set up your package structure.
2. In your package, create one or several modules where you define your tools using the @tool decorator.
3. You can include your own user-configurations, and load them using the get_config_value function from the toolit.config module.
4. Make sure to include toolit as a dependency in your package's setup.py or pyproject.toml.
5. Register your plugin with Toolit by adding an entry point in your setup.py or pyproject.toml, so Toolit can discover your tools when the package is installed. The entry point is called toolit_plugins.
6. Publish your package to PyPI or install it from a git repository where you need it.

See an example plugin here: toolit-azure-devops-trunk-based-branching

Contributing

We welcome contributions to Toolit! If you have ideas for new features, improvements, or bug fixes, please open an issue or submit a pull request on our GitHub repository. We appreciate your feedback and support in making Toolit even better for the community.