fix(studio): recover transient AI failures and report provider diagnostics

README.md

@@ -141,7 +141,7 @@ AI generation sends the current design, brief and allowed component capabilities
 to APIMart when you request generation. The engine validates proposals and locks
 before offering them for review. Failed requests leave the draft unchanged.
 Requests allow two concurrent generations and ten starts per minute per server
-process, with two attempts, 4096 output tokens per attempt and a 60-second total deadline.
+process, with two total upstream requests (including repairs and temporary-failure retries), 8192 output tokens per attempt and a 60-second total deadline.
 Without APIMART_API_KEY, generation uses local rules and reports that in the UI;
 local rules do not interpret arbitrary product descriptions. Drafts remain in
 browser storage; export before clearing browser data or moving to another origin.