ClickUp's depth is its selling point in the app and its main problem over MCP. Workspace, space, folder, list, task, subtask: to act on one task an assistant may need several calls just to find it. Every setup that feels slow is a scoping problem, not a performance problem.
Scope first, then ask
Give the assistant the list you mean. "In the Marketing space, Q4 folder, Launch list" turns a tree walk into a single lookup. If you work mostly in two or three lists, keep their IDs somewhere you can paste; it changes the experience completely.
Custom fields are the honest weak spot
Teams that use ClickUp seriously live in custom fields, and community MCP servers implement them unevenly. Plenty can read a custom field and cannot write one. Before building any routine on top of this, test the exact write you care about — do not infer support from the README.
What it is good at
Intake and triage. A paragraph becomes a correctly placed task with an assignee and a due date. Cleanup, too: finding tasks with no assignee or no due date across one list is a question ClickUp's own filters answer, but it is faster to ask, and the assistant can fix them in the same breath.
Where LemTask stands today
LemTask does not ship an MCP server yet, and we are not going to pretend otherwise on a page developers will read. What LemTask has working right now is outbound and inbound webhooks, Zapier, Make and Apple Shortcuts, which covers "when a task is created, do X" and "create a task from anything that can send an HTTP request". An MCP connection is on the roadmap; when it ships and we have tested it end to end with Claude, this page will say so with the config block, and not a day earlier.