Sharing
Sombra lets you share collections in two ways: publicly with anyone via a link, or privately across workspaces in your account.
Public sharing
To share a collection publicly, open the collection and click the gear icon to access settings. Under Public sharing, you have two options:
Share as snapshot
Creates a static, point-in-time copy of the collection. The public link shows the collection as it was when you shared it. If you add or remove artifacts later, the snapshot doesn't change. Good for sharing a finished set of research or a curated reading list.
Share as live
Creates a public link that always shows the current state of the collection. As you add, remove, or update artifacts, the public view updates too. Good for ongoing research that others want to follow.
Both options generate a unique share URL (e.g. sombra.so/s/{id}) that anyone can view — no Sombra account required.
Workspace sharing
Collections can also be shared across workspaces within your account. This is useful when you have separate workspaces for different projects or teams but want to reference the same collection in multiple places.
From the collection settings menu, the Workspace sharing section shows which workspaces the collection is currently shared with. Shared collections appear under "Shared with you" in the collection list of the receiving workspace, tagged with the source workspace name.
Shared collections are read-only in the receiving workspace — only the owner can add, edit, or remove artifacts.
Public Collections page
The Public Collections page in the sidebar shows all collections you've shared publicly. From here you can see what's live, review what you've shared, and manage your public links.
MCP and sharing
Collections shared to a workspace are accessible via MCP tools too. The browse_collections tool returns shared collections alongside owned ones (with a permission field so agents know what's read-only), and read_collection works on shared collections just like owned ones.
This means your AI assistant can read reference collections from other workspaces — for example, pulling in a shared library research collection as context while working in a different project workspace.