lovethedocs

Polish your Python docs in 30 seconds.

Make your docstrings excellent with a single command. It’s fast (~30s), cheap (~10¢ for small repos), and non-destructive.

• One‑command upgrades – lovethedocs update path/
• Choose your style – NumPy (default) or Google with -s/--style
• Safe by defualt – edits live in path/.lovethedocs/ until you accept them
• Parallel & fast – set -c/--concurrency to speed things up

Quick start

pip install lovethedocs

# 1. Get an OpenAI API key (https://platform.openai.com) and set it:
export OPENAI_API_KEY=sk-...

# 2. Stage docstrings for your project (NumPy style):
lovethedocs update path/to/project/

# ...or choose Google style and review in one shot with 8 workers:
lovethedocs update -s google -r -c 8 path/to/project/

Review & clean

lovethedocs review path/to/project/   # open diffs in your viewer
lovethedocs clean  path/to/project/   # wipe staged edits

Tested on macOS; supported diff viewers: cursor, code, git, terminal. Help adding others is welcome!

🎯 Example

Before:

def process_data(data, threshold):
    # Process data according to threshold
    result = []
    for item in data:
        if item > threshold:
            result.append(item * 2)
    return result

After:

def process_data(data: list, threshold: float) -> list:
    """
    Filter and transform data based on a threshold value.

    Parameters
    ----------
    data : list
        The input data list to process.
    threshold : float
        Values above this threshold will be processed.

    Returns
    -------
    list
        A new list containing doubled values of items that exceeded the threshold.
    """
    result = []
    for item in data:
        if item > threshold:
            result.append(item * 2)
    return result

CLI cheatsheet

Contributors welcome

Issues, discussions, and PRs are encouraged—especially new doc styles, diff viewers, or provider integrations. Open an issue to get started.

lovethedocs is free to install; LLM calls are billed by OpenAI at your usage rates.

📄 License

This project is licensed under the Apache 2.0 License - see the LICENSE file for details.