Preset Feature Feedback
What is the problem or goal you're trying to solve or accomplish?
When using Preset MCP to query a huge dashboard with more than 100 tabs and 100 charts it returns a truncated chart list.
How are you solving it currently?
Not possible
What is your recommended solution?
Add a list_dashboard_tabs(dashboard_id) tool and an optional tab_id filter to get_dashboard_layout, returning only that tab's subtree.
Or equivalent server-side scoped or paginated layout discovery would also work.
Or make the response limit configurable or large enough for large dashboards.
Thanks for opening this request. This currently blocks our MCP-only reporting workflow on dashboard 51 (114 tabs, 404 charts). get_dashboard_layout retrieves the layout but returns “Response too large: ~59,402 tokens (limit: 20,000)” instead of delivering it. get_dashboard_info returns only 100 of the 404 charts and does not provide the tab hierarchy, so our agent cannot reliably identify the correct tab for a report-row permalink.
As a short-term mitigation, could the server-side response limit for get_dashboard_layout be raised or made configurable to at least 80,000 tokens, preferably as an opt-in setting for this tool or workspace? That should accommodate the currently measured response if no other limit applies. We can control when our agent requests the full layout and how it processes the result.
Could you confirm whether this mitigation is feasible, whether another response limit would still apply, and whether it could be delivered sooner than tab-scoped discovery? We still support list_dashboard_tabs and a tab_id-scoped layout response as the scalable long-term solution.