Context
When we create a knowledge base on Kaapi, the API currently returns a top-level success/failure response. It does not tell us which specific files failed and why they failed (e.g. unsupported extension, parse error, size limit).
Problem
We have a use case (cloning a legacy assistant) where we upload a batch of files to Kaapi and create a new knowledge base. If any file in the batch is rejected during knowledge base creating, we currently have no programmatic way to know which ones failed and why — we only know the overall request failed. This forces us to either pre-filter aggressively on our side or treat the entire clone as a failure.
Request
Please include failed file details in the API response for knowledge base creation / document upload. Specifically:
- Which files were rejected (by file ID, name, or some identifier)
- The reason for each rejection (e.g. unsupported_format, file_too_large, parse_error)
Why this matters
With this information, Glific can:
- Decide whether to proceed with a partial clone or abort
- Surface meaningful errors to org admins instead of a silent failure
Context
When we create a knowledge base on Kaapi, the API currently returns a top-level success/failure response. It does not tell us which specific files failed and why they failed (e.g. unsupported extension, parse error, size limit).
Problem
We have a use case (cloning a legacy assistant) where we upload a batch of files to Kaapi and create a new knowledge base. If any file in the batch is rejected during knowledge base creating, we currently have no programmatic way to know which ones failed and why — we only know the overall request failed. This forces us to either pre-filter aggressively on our side or treat the entire clone as a failure.
Request
Please include failed file details in the API response for knowledge base creation / document upload. Specifically:
Why this matters
With this information, Glific can: