🛠️ How-to guide · Last updated: October 1, 2026

<aside> 📄

WHAT IT DOES

Sources is where you add everything your agents learn from — documents, websites, manual notes, cloud files, SharePoint libraries, database records, and content from your own systems' APIs — and organize it with tags. It's the single list of all your content, whichever knowledge base it ends up in.

</aside>

What the Sources page shows

The Sources page lists every piece of content in your workspace, with its title, source type, and when it was last updated. Tabs across the top group content by processing status — All, Ready, Processing, Failed — each with a live count. Search across everything, filter by source type or tag, group by a tag, and use Manage Tags to organize your labels.

The Sources list — every piece of content in the workspace with its source type, processing status, and last update, plus search, filters, and tags.

The Sources list — every piece of content in the workspace with its source type, processing status, and last update, plus search, filters, and tags.

Adding a source

Select Add source and pick where the content comes from. Each choice opens a short guided wizard in the same window; the final step ingests the content and returns you to the list.

The Add content picker now offers eight source types. The two new ones, API Crawl (1) and API Push (2), bring in content from your own systems.

The Add content picker now offers eight source types. The two new ones, API Crawl (1) and API Push (2), bring in content from your own systems.

The source types

You can add content from eight kinds of source:

  1. Manual entry — Compose an entry yourself. Give it a title, choose a format (Markdown, Plain text, or HTML), write the content, and add tags. Good for a short notice like Riverton's holiday collection schedule.
  2. Upload files — Upload one or more documents by drag-and-drop. Supported types include PDF, Word, PowerPoint, Excel, text, Markdown, HTML, CSV, JSON, and more. Good for the Waste Collection Schedule PDF or Permit Guidelines.
  3. Website crawl — Crawl a public website or help center and keep it fresh on a schedule. See Crawls & Syncs for the full setup. Good for crawling riverton.gov.
  4. Cloud storage — Sync from a cloud storage container. Name the source, set a refresh frequency (Hourly, Daily, Weekly, or Monthly), then add your files to the container Zammo provides and ingest. Good for a shared folder of city forms.
  5. SharePoint — Connect to SharePoint and sync one or more document libraries. Sign in, choose the libraries, and finish. Good for internal policy libraries.
  6. SQL Query — Run a read-only query against an external database and ingest the rows as content. Set a name, the connection details, the query, and which columns hold the row ID, title, and body. Good for a live list of open service requests.
  7. API Crawl — Pull records from your own system's REST API on a schedule. Import the API's OpenAPI description (or enter the endpoint by hand), connect the fields that matter, and Zammo keeps the records in sync. Good for the Parks & Recreation facilities list.
  8. API Push — Let your own systems send content to Zammo through the push API whenever it changes, using a workspace API key. Pushing the same record again updates it. Good for road closures from the Public Works work-order system.

Cloud storage, SharePoint, SQL Query, and API Crawl are available on paid plans — on other plans the tile shows an upgrade prompt. Manual entry, Upload files, Website crawl, and API Push are on every plan; the API keys for pushing are generated by workspace Admins.

Full step-by-step guide for each: Add a source: Manual entry · Add a source: Upload files · Add a source: Website crawl · Add a source: Cloud storage · Add a source: SharePoint · Add a source: SQL Query · Add a source: API Crawl · Add a source: API Push