Collect Your First Source
Start with a working instance and a connected Web app or Agent. The guides default to Web instructions, with a CLI / Agent alternative. You do not need to connect every account at once: choose one small source with an item you will recognize.
Choose your first source
RSS or Atom is a useful first choice because a public feed needs no account credentials. If you prefer your own saved information, start with GitHub Stars, email, Telegram, or Twitter / X bookmarks. Each guide includes its own prerequisites and setup.
An Extension supplies a collector implementation; a Source is one configured use of it. For example, install the RSS Extension once, then create one Source per feed. Installing it does not automatically enable it or collect anything. Prepare the Extension on Core, then configure the Source in the Web app. Select the online Core client when installing a Python-only collector; the browser does not need to run that Extension itself.
Complete the first loop
- Follow your chosen source's guide to enable its Extension and create the Source. Open its details in the Web app, or retain the returned Source ID when using the CLI.
- Run a Collection and wait for the returned Job ID to finish. These are different IDs: the Source persists across runs; each Job represents one run.
- Find What You Saved: maintain the lexical index and search for a known item.
- Only after that works, schedule collection and indexing.
Done means you retrieved a real item, not just that installation or a Job succeeded. A successful empty collection may be normal; each source guide explains what is eligible for collection.
Next: Connect More Sources, or Use Your Information.