When documents contain embedded hyperlinks (e.g., text like "click here" linking to a URL), Glean Chat, Agents, and API/MCP only return the visible text - not the actual URL behind it. Users are told to open the source document and click the link manually.
This is a poor experience, especially when:
- Documentation is full of embedded links (Confluence, SharePoint, Google Docs, Word, PDFs)
- Users ask Agents to extract all links from a document
- Printing full raw URLs in docs is not a realistic workaround
Hyperlink extraction already works in the /search REST API. We'd love to see this extended to Assistant, Agents, /chat API, and MCP so links are served directly in responses.
Upvote if your team would benefit from this!
