Create a Discovery without leaving chat, and keep working with finished Discoveries by attaching them as context.
Discoveries don’t have to start on the Discovery page. From any chat you can spin up a new one, watch it fill in inline, and, once it’s done, keep asking follow-up questions with the Discovery attached as context.
Discovery citations resolve at three levels: column, row, or cell, so “why is row 3’s job title Innovation Manager?” lands on that cell.
Either pick Tools → Create new → Discovery explicitly, or just describe what you want and Libra figures out which tool to use.
1
Pick + Tools → Create new → Discovery (or describe what you want)
A Discovery badge pins under the chat composer, confirming what’s about to be created.
2
Attach the documents and describe the columns
Drop in the documents you want to extract from. Type a prompt that names the fields, like “Extract employee name, job title, and home office location for each contract”. Libra builds a Discovery with those columns and runs it across the attached documents.
3
Send
Libra creates the Discovery, runs the extraction, and shows the result inline as a preview card. Each row corresponds to one document; columns appear as Libra extracts them.
4
Open in Discovery to refine
Click Show in Discovery to open the full Discovery view. Add or remove columns; the Discovery re-runs immediately for any new column.
Discoveries created from chat land in the same project as the chat. They’re saveable as templates from the full-screen view; the second time you build something similar, save it for next time.
Once a Discovery exists, attach it to any chat as context and ask follow-up questions. Libra answers with citations to specific columns, rows, or cells.
With @-mention while typing
Type @ anywhere in the chat input. The picker opens grouped by type. Pick the Discovery. The Discovery is now part of the chat’s context, and Libra can cite it in answers.
Via Tools → Add context
Click + Tools above the chat input and choose Add context. Pick the Discovery from the picker. Same effect.
When you attach a project as context (rather than a single Discovery), every Discovery in that project becomes referenceable. Useful when you want Libra to consider all the Discoveries on a matter.
Discoveries are project-scoped. The chat needs to be in the same project (or you can attach the project itself).
2
Type @ and pick the Discovery
The picker shows recent items grouped by type. Discoveries appear with the Discovery icon.
3
Ask your question
Anything from filter to aggregation to a free-form summary.
4
Read the answer with cell-level citations
Inline citations link to specific cells in the Discovery. Open the Cited Sources panel and the Discovery card expands to a mini table preview with the cited cell highlighted, so you can verify the cell context without opening the full Discovery view.
Attach the documents before sending (creation flow)
Discoveries need documents to work on. Without attachments, Libra can’t extract anything; it’ll fall back to a regular chat answer.
Be specific about what you want
“Summarise this Discovery” gets generic. “Summarise the high-value contracts in row 1, 4, and 7, emphasising non-compete duration” gets something usable.
Ask for citations back
“And cite the relevant cells” on any prompt gets you a verifiable answer.
Stack context if you need it
Attach the Discovery and a Review and the underlying documents. Libra can cite into all of them; “explain why Topic 3 in the Review is medium risk and which Discovery rows are most affected” is a perfectly usable prompt.
Don't re-extract data in chat
If you find yourself asking “now extract the notice period for each contract” in a chat, run a column update on the Discovery instead. Discoveries are deterministic; chat reasoning isn’t.