Knowledge

Knowledge gives your agent searchable context from your own content — product documentation, company policies, codebases, wikis — for accurate, grounded answers.

Tip

Only need the agent to read a file for a single task? You don’t need a collection — just attach it to the message. Reach for Knowledge when the same content should be reusable and searchable across many chats.

Create Knowledge

Set up a knowledge collection and upload or connect your content. Create →

Connect Data Sources

Sync from Notion, Confluence, Google Drive, GitHub, and more. Sources →

Sync & Manage

Keep your knowledge up to date and monitor sync status. Manage →

---

Supported sources

Source	Type
File upload	PDF, Word, TXT, Markdown, CSV
Notion	Pages and databases
Confluence	Spaces and pages
Google Drive	Docs, PDFs, folders
GitHub	Repos (docs, README, source)
Web Crawler	Any public website

---

How knowledge retrieval works

When you attach knowledge to an agent:

1. The user’s message is turned into a semantic search query
2. Auteryn finds the most relevant chunks from your knowledge collection
3. Those chunks are passed to the agent as context
4. The agent answers based on your content — not just its general training

This means your agent can answer questions specific to your product, your team, or your data — things a general AI would have no way of knowing.