My API endpoints probably work when everything goes right. I want to know what happens when things go wrong. For every API route or server endpoint in my app, write tests that cover the error paths: 1. Requests with missing required fields. 2. Requests with the wrong types (strings where numbers belong, arrays where objects belong). 3. Requests with no authentication token where one is required. 4. Requests with a valid token for a user who shouldn't have access to that resource. 5. Requests with oversized payloads or absurdly long strings. 6. Requests for resources that don't exist (bad IDs, deleted records). For each case, assert two things: the response status code is correct (400, 401, 403, 404 as appropriate) and the response body does not leak a stack trace, internal file paths, SQL errors, or other implementation details. Don't change the endpoint implementations unless a test exposes a genuine crash or information leak. If you find one, stop and show me what you found before fixing it. When finished, run the full suite and give me a summary: which endpoints handled bad input cleanly and which ones returned 500s, stack traces, or unexpected data.