Currently, when a user imports a knowledge source via URL into Respond.io, the system automatically crawls up to three levels of child pages. Although this enriches the AI's training material, workspace managers or owners might need more control. They may wish to precisely determine which child pages should be included or excluded, ensuring that the AI focuses on the relevant content and minimizes the potential for conflicting information.

Introduce a feature that enables managers or owners of the workspace to selectively exclude certain child pages from being used by Respond AI. This enhancement will optimize the efficiency of Respond AI, reduce the likelihood of errors and internal conflicts, and create a more conducive environment for new agents.

Currently there is no workaround for this problem as all child pages up to 3 levels will be included as a knowledge source for Respond AI.