Privacy Policy

Last updated: September 4, 2026

LegacyDoc turns source code into documentation using Claude, Anthropic's AI model. This page explains exactly what we collect, what we send to a third party to do that, and what we never keep.

The short version: the code you upload or select from GitHub is sent to Anthropic's Claude API for a single documentation request, then discarded. We don't store it, log its contents, or use it to train any model.

1. What we collect

Depending on how you use LegacyDoc, we handle:

Uploaded files or pasted code — held in memory only for the duration of your documentation request.
GitHub account data, if you connect GitHub — your username, avatar, and a list of repositories you have access to, plus the contents of the specific repository you choose to document.
A GitHub access token, scoped to read your repositories, issued when you authorize the app and used only to fetch the repository you select.
Basic request metadata — timestamps, response status, and error codes, captured in standard AWS request logs for debugging. These logs do not contain your code or file contents.

We do not require an account, a password, or an email address to use LegacyDoc directly.

2. What we send to Claude (Anthropic)

Documentation generation is powered by Anthropic's Claude API. When you click Generate Documentation, the code you provided is sent to Claude in a single API request, and the response is returned to you as your documentation. No other party sees this content.

Under Anthropic's standard commercial API terms, prompts and outputs sent through the API are not used to train Anthropic's models and are retained by Anthropic only briefly for abuse and safety monitoring, not for any other purpose. See Anthropic's own privacy policy for their side of this. If our data-processing terms with Anthropic change, we'll update this policy.

3. What we don't do

We don't persist your uploaded code or repository contents to a database or disk after your request completes.
We don't use your code to train, fine-tune, or evaluate any model, ours or anyone else's.
We don't sell, rent, or share your code or GitHub data with advertisers or data brokers.
We don't currently use analytics, advertising, or tracking cookies. If that changes, this policy will be updated first.

4. GitHub integration

When you connect GitHub, LegacyDoc requests read-only access to your repositories so it can pull the code for the repo you select. You can revoke this access at any time from LegacyDoc (the Disconnect button) or directly from your GitHub application settings.

5. Data retention

Uploaded code and repository contents exist only for the lifetime of a single request and are not written to persistent storage. GitHub access tokens are held only for your session and are discarded when you disconnect. Infrastructure request logs (timestamps, status codes) are retained for a limited period for debugging and abuse prevention, then deleted.

6. Your rights

Because we don't retain your code or require an account, there's generally nothing of yours stored to access, export, or delete. If you connected GitHub and want your access revoked, use the Disconnect button or your GitHub settings. For any other privacy question or request, contact us below.

7. Changes to this policy

If this policy changes in a way that affects how your data is handled, we'll update the "last updated" date above and, for material changes, note it on the homepage.

8. Contact

Questions about this policy or how your data is handled: kareem.rufai27@gmail.com.