Code documentation your team can actually trust.

No. Your code is analyzed in memory and never written to disk. We use GitHub OAuth to read your repository, generate documentation, then discard the source. Only the generated documentation is stored.

JavaScript, TypeScript, and Python have full AST parsing via Tree-sitter, which powers the deepest documentation and code navigation. Other languages get best-effort support: Claude AI can still read and describe source files, but without the structured AST analysis that JS/TS/Python get.

We register a GitHub webhook on your repository. When you push code, GitHub notifies AutomaDocs and we regenerate documentation for changed files. The process typically completes in under 2 minutes.

Yes. All plans support private repositories. We use GitHub OAuth with the minimum required scopes - read access to your code and permission to register webhooks.

If you are not satisfied within 14 days of your first paid subscription, we will issue a full refund. No questions asked.

You ask in plain language; we search your generated docs and code context, then answer with links to files and line numbers. Under the hood we combine keyword and semantic search (hybrid retrieval) so answers stay grounded in your repo.

No. AutomaDocs is self-serve: documentation, FAQ, and status first. For bugs or billing, email hello@automadocs.com. We reply when we can, usually within a few business days. There is no 24/7 phone line, SLA, or account manager.