mask4llm 1.0.0

A terminal user interface (TUI) app built with Textual to easily mask and unmask sensitive information.

mask4llm

A terminal user interface (TUI) tool to safely mask and unmask sensitive data before interacting with large language models (LLMs). Built with Textual, it lets you obfuscate protected fields in your data before sending to an LLM, and then restore the originals after receiving the LLM's response.

Features

• Mask sensitive info (names, emails, etc.) before sending prompts to LLMs.
• Unmask LLM output swap masked placeholders back after processing.
• Modern TUI built with Textual, with easy navigation.
• Clipboard support via pyperclip.
• Simple installation, cross-platform support.

Installation

Install via pip (PyPI):

pip install mask4llm

Or, if using the uv package manager:

uv pip install mask4llm

Once installed, launch the app with:

mask4llm

Development Setup

This project is managed with uv and uses Just for dev tasks. The dev dependencies include basedpyright, ruff, pytest, and pre-commit.

1. Clone the repository and install Just and uv

git clone https://github.com/hemanthhari2000/mask4llm.git
cd mask4llm
# Install uv and just using your OS/package manager

2. Sync the Development Environment

just install
# or, directly:
uv sync

This sets up .venv and installs all runtime and [dev] dependencies.

3. Run App in Development

uv run mask4llm

Common Just Commands

Command	Description
just install	Sync env & install all dependencies
just update	Upgrade all dependencies
just lint	Run Ruff for linting (with autofix)
just format	Run Ruff for formatting
just typing	Type-check with basedpyright
just test	Run all tests with pytest
just pre-commit	Run pre-commit hooks
just clean	Complete cleanup (caches, pyc files, venv)
just fresh	Clean and reinstall all dependencies
just all-dev	Run lint, format, and type-check (fast dev check)

Usage/Examples

• Mask Data:

  • Paste or type sensitive data, run mask, and copy masked output to LLM prompt.

• Unmask Output:

  • Paste LLM output, run unmask, and restore sensitive data.

Details and instructions are shown in the TUI footer and help screens.

Contributing

Contributions are welcome! Please open issues or PRs as needed. Run all checks before submitting:

just all-dev && just test

License

This project is licensed under the MIT license. See LICENSE file for details.

Acknowledgements

• Textual for the TUI framework.
• uv, ruff, and basedpyright for the awesome developer tools.