There are 3 updates, 1 comment.
 
 
LLM AI Integration / cid:jira-generated-image-avatar-8fda124c-3f6e-40ab-98bb-2a00e39f92bc LLMAI-190 Closed

Plain rendering glues message-box titles and adjacent text together

 
View issue   ยท   Add comment
 

3 updates

 
cid:jira-generated-image-avatar-96292a99-e876-4ee6-bc5f-7cd1c385982e Changes by Paul Pantiru on 06/Nov/26 11:10
 
Assignee: Paul Pantiru
Resolution: Won't Fix
Status: Open Closed
 
 

1 comment

 
cid:jira-generated-image-avatar-96292a99-e876-4ee6-bc5f-7cd1c385982e Paul Pantiru on 06/Nov/26 11:10
 

Closing as Won't Fix. The glued words come from the platform's plain/1.0 renderer, which prints no separator before inline content placed directly in a group (XRENDERING-832); any page shows it with ?outputSyntax=plain. Avoiding it in the MCP server would mean rendering the executed page ourselves instead of calling displayDocument, which is not worth it for the plain mode: agents cope with the glued words, and format="html" keeps the structure of message boxes and images.