Show HN: I built a GitHub AI app to automate the fixing of outdated docs

In my 15+ years as a developer, one of the most annoying things I've consistently faced is outdated documentation. I am also not a saint either, I skip updating docs myself. So I started working on tools to make maintaining docs a bit easier for developers like us. There have been many attempts to solve the problem of continuous documentation. Tools like Swagger or Sphinx-autodoc can generate docs from OpenAPI specs or code annotations, but they only work in very narrow settings. They don't help much with higher-level docs like user guides, quickstarts, or example scripts. And they still rely on developers to manually maintain intermediate elements like docstrings. With LLMs, updating high-level docs has gotten easier. You can prompt your AI IDE (e.g. Cursor) to update docs, and it usually works. But this approach has a few issues: - You have to remember to prompt after every code change. - You don't know exactly which files are being used in context—either you specify them manually or