The page was making 1 API call per tree to fetch session stats, but the
backend sessions endpoint doesn't support tree_id filtering — so every
call returned identical data. Now fetches trees and sessions in parallel
(2 calls total), builds a lastUsed map client-side, and uses the existing
usage_count field from the tree list response for session counts.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
0c9fbdc90b