Errors will happen after launch, and I want to know my app fails well: users see something helpful, I see something actionable, and nobody sees a stack trace. Audit the error handling across my app: 1. Server errors: when an endpoint throws, what does the client receive? Map which endpoints return clean error responses (correct status, useful message) versus raw error objects, stack traces, or HTML error pages where JSON is expected. 2. Client errors: when an API call fails in the browser, what does the user see? Find the places that show a useful message, the places that show nothing (silent failure), and the places that show raw technical output. 3. Unhandled paths: look for promise rejections without catch handlers, async functions without error handling, and event handlers that can throw and take down the UI. 4. Logging: when errors happen server-side, where do they go? Verify errors are captured with enough context to debug (which endpoint, which user, what input class) but without logging secrets, passwords, tokens, or full payment details. 5. The 404 and 500 pages: confirm they exist, look intentional, and don't expose framework version banners or internal paths. Deliver a findings list organized by severity: information leaks and silent failures first, cosmetic issues last. Include the file and line for each. Don't fix anything yet. I want the audit first, then I'll tell you which findings to fix.