As a user, I want to search for documents, conversations, text blocks, and media so that I can quickly find and navigate to the information I need across my knowledge base.

Concept 01 → Search

Create simpler structures for search, navigation, and information design to improve findability, accessibility and usability.

PRD: Explore Knowledge

Design

Browser

First sketches hm://z6MkrbYsRzKb1VABdvhsDSAk6JK8fAszKsyHhcaZigYeWCou/:profile?v=bafy2bzacecoevvu2ebpntbrhm552uqp3ag5vrxvukpk7a4vvfmhazkk7smu5y&l, concepts overview for browser.

We can add a "spaces filter" if we want to give the option to outside the space user is in.

Desktop

For desktop would be similar experience. Search icon need to be on the omni bar (pretty important to understand you can search). Need to see how we incorporate Agent search.

Ask agent concept

Have a conversation with the data. The Agent needs to fit naturally into the user's actual workflow, the job they're trying to get done (in this case search).

Agents should be a tool to manage and research content.

Acceptance criteria:

The search input is prominently placed and visually distinct, larger in size, with placeholder text and an icon, so users can immediately locate it without scanning the navigation.

In browser the search is limited to space user is in

User clicks on search area, a dropdown is open with recent searches and option to switch to "Ask agent" for a more conversational search.

Wnen user starts typing any word a mix of results appear with maximum number of results (TBD) and then a button to "show all results".

"Show all results" opens a full screen page with a long list with all results mix and the number per all types (documents, text blocks...) and the option to filter by Type (docs, text blocks...), In (section/navigation item), Author (who created it) and publication date

"Add filter" option opens a list with all filters

User goes back to main page with back arrow

Concept 02 → Explore (full screen)

Working on now, how we can merge the concept of search and explore multiple nodes and types of content and filters.

Query builder aplicado a búsqueda buscar ejemplos hm://zDnaenBn1Xv8hiFC3dcgikrjk4UzurF1aeqVWQ1TNnnLG34qo/:profile?v=bafyreigreppmjj5mwcbws2gspmjcv3kefcr3d3m3p5hnqhbaw53arav5ey&l pásame casos de uso