Skip to content
SecondStack SecondStack SecondStack Docs

Knowledge Collections

Knowledge Collections in SecondStack let you organize documents and use them as grounded context in AI chats. A collection is a named, user-owned set of files that exists independently of any single chat and can be referenced from any thread. Files are automatically indexed for retrieval, so models answer from your content instead of guessing.

Full transcript lightly edited for readability

Knowledge Collections overview (00:00). Hi everyone. Today I’d like to tell you about the Knowledge Collections feature in SecondStack. What is it? Knowledge Collections lets you store files and client materials as a collection and reuse them as context in different threads. Usually, if you want the model to use some documents, you have to upload those files again in every new thread. With Knowledge Collections, you upload them once and reuse them later as an attachment.

Creating a collection (00:45). Let me show you how it works. First, we create a collection. Then I upload my test files, and I can use drag and drop here. Once the files are processed, they become available as reusable context in different threads.

The Knowledge Collections page (01:10). Before I show you how it works in a chat, a quick look at this page. Here we can manage collections and the files inside each collection. For every file I can change the auto-generated comment, write my own, or regenerate it. These comments are visible to the model and help it understand your files better.

Using a collection in a chat (01:45). To use a collection, I click the attach button and select my collection. Now I’ll ask a question about the files: can you see the docs? We get the answer, and as you can see, the model sees all the files stored in the knowledge collection and attached to the message. It reads the files and answers based on them. If you want to use this collection again, you create a new chat and attach it to a message there.

Summary (02:40). That’s the whole idea: you create a collection once, add all the necessary files, and reuse it in any chat you want. If you have any questions or comments about this video, please leave a comment below. Thank you for watching.

System Design

File ↑ Comment (model-visible) Mode Tokens
architecture-backend.md — RAG 3k
architecture-frontend.md — RAG 4k
sync-jobs.md Background Sync Jobs RAG 1k

A collection groups related files; each is indexed for retrieval and referenced by name in chat.

How it works

Section titled “How it works”
  • Create a collection with a name and description, and add files to it.
  • Files are auto-extracted and embedded into a vector database when added.
  • In a chat, reference a collection (via a # trigger or the toolbar) to make its contents retrievable.
  • At response time, the model retrieves the most relevant passages from the collection to ground its answer. Presets can also inject a document’s full content when full-context grounding is preferred over retrieval.

Sharing and permissions

Section titled “Sharing and permissions”

Collections are shareable with fine-grained access. Each grant carries a permission level:

  • Read — use the collection in chats.
  • Edit — add or remove files.
  • Manage — control sharing and settings.

You can share with individual users, groups, or teams, and team and group membership is resolved live from ControlTower. An organization-wide grant is capped to read access.

Helpful touches

Section titled “Helpful touches”
  • Auto-generated file descriptions give models a short, useful summary of each file.
  • Cascade delete removes a collection’s files, stored blobs, and index entries together, so nothing is left behind.
Section titled “Related”