Version 3.26 of the Python feed reader library, reader, has been released. A significant new feature is automatic feed discovery, where users can input a website URL and the system will suggest found feeds. This functionality is powered by the autodiscover plugin, which stores discovered feeds in a feed tag. Additionally, reader now supports database exports, allowing users to download their entire data, including feeds and article statuses. This feature is particularly useful for migrating to other reader installations or for local backups. The developer is also working on a hosted version of reader, with a public, read-only demo available for exploration. Key completed features for the hosted version include a multi-user web app, email authentication, and infrastructure deployments. The reader library itself provides core feed reading functionalities such as retrieving, storing, and managing various feed types, marking articles, tagging, filtering, searching, and obtaining statistics. It offers a stable API, extensive test coverage, and typed Python. The library aims to simplify the process for developers who want to build their own feed readers or enhance existing ones. reader is designed to be a long-term, extensible, stable, and well-tested library with minimal dependencies.