Is your feature request related to a problem?
The quality of AI-generated code can vary. Currently, if a user receives a translation that is incorrect, suboptimal, or buggy, they have no easy way to report it. Without this feedback, it's difficult for us to identify weaknesses and improve the core AI prompts.
Describe the solution you'd like
-
In the injected UI, next to the "Copy" button, add a simple "👍" (Good) and "👎" (Bad) icon button.
-
When a user clicks "👎", open a small, simple modal with a textarea, asking "What was wrong with this translation?".
-
The feedback (including the original code, the bad translation, the target language, and the user's optional comment) should be sent to a new, dedicated backend endpoint.
-
This feedback can be stored in a database (like Supabase/Firebase).
Additional context
Creating a direct feedback loop is one of the most effective ways to gather data for improving the quality of the AI. This feature would provide invaluable insights into where the model struggles.
Is your feature request related to a problem?
The quality of AI-generated code can vary. Currently, if a user receives a translation that is incorrect, suboptimal, or buggy, they have no easy way to report it. Without this feedback, it's difficult for us to identify weaknesses and improve the core AI prompts.
Describe the solution you'd like
In the injected UI, next to the "Copy" button, add a simple "👍" (Good) and "👎" (Bad) icon button.
When a user clicks "👎", open a small, simple modal with a textarea, asking "What was wrong with this translation?".
The feedback (including the original code, the bad translation, the target language, and the user's optional comment) should be sent to a new, dedicated backend endpoint.
This feedback can be stored in a database (like Supabase/Firebase).
Additional context
Creating a direct feedback loop is one of the most effective ways to gather data for improving the quality of the AI. This feature would provide invaluable insights into where the model struggles.