Sloth-ninja closes a key-management reliability and privacy gap
This fork makes API-key administration less brittle while reducing the risk that backend failures reveal too much.
Sloth-ninja has fixed a crash when checking the status of user API keys, a small but consequential repair for any team relying on external services inside a legal-AI workflow.
The update also makes genuine key-saving failures visible to the person operating the system, while tightening what a server error sends back. Unhandled failures now stop the affected process instead of leaving it stuck in a broken state. A related test has been given more time to complete where secure password-style hashing can run unpredictably slowly.
Spotted something wrong? Or know the PR text has fresher detail than the writeup above?