Skip to main content

workspace_mismatch

Emitted on the MCP surface when a tool argument carries a workspaceId that differs from the connection’s active workspace. The fix is in the message: call switch_workspace to change the active workspace first, then retry. On REST, the analogous failure is workspace_unavailable (a Scripe-Workspace-Id header the token cannot honour), and plain cross-workspace resource probes return not_found — the API never confirms the existence of another workspace’s resources.