Rodolphe
We use the ClickUp MCP server with Claude to answer project questions, for example checking tracked time against client quotes. Tracked time is currently hard to retrieve through the MCP server, and the workarounds put unnecessary load on your API.
What we observed
  • clickup_get_time_entries only filters by task_id, dates, assignee and billable status. There is no list_id, folder_id or space_id filter, although the REST endpoint GET /api/v2/team/{team_id}/time_entries supports location filters such as list_id.
  • Without a task_id, results are capped at 50 entries per call, and the tool exposes no pagination parameter, so totals are silently truncated.
  • time_spent is not returned by clickup_filter_tasks, and it comes back null from the bulk task fetch (task.get_many) even for tasks that have tracked time.
Impact
To check one quarter of tracked time for three client folders (over 100 tasks), we had to query tasks one by one 🐌 : more than 60 separate requests for a single report, where one request per list would have been enough :)
Request
  • Add list_id / folder_id / space_id filters to the time entries tool, with pagination (cursor or page) so results are never silently truncated.
  • Return time_spent (and time_estimate) in task list results and in the bulk task operator.
  • Optionally, offer an aggregated mode (total per task, list or user) so individual entries do not need to be transferred.
This would significantly reduce the number of calls hitting your API and make time-tracking reports reliable for MCP users.