Fix 403/429 Node API Errors

My Amazon-Hiring job grabber is built with Node.js and already lives on its own server. I run several concurrent logins, then cycle through three modes that hit Amazon’s internal APIs. Right now that flow breaks:

• automatic sign-in returns 403 not everytime
• Mode 0 works briefly, then 403s; every request here must come back in under 100 ms with zero jitter
• Modes 2 & 3 intermittently throw both 429 and 403; each call has to stay under 500 ms, again without jitter

I need you to pinpoint why those status codes appear, reinforce the authentication flow, and tune the request strategy so the grabber can run nonstop across all three modes. That includes:

– redesigning or refactoring the Node.js call sequence for minimal latency
– applying solid rate-limit handling, header tuning, cookie / token refresh or any proxy logic you deem necessary
– keeping CPU and memory footprints low so the server never becomes the bottleneck

Acceptance criteria
1. Sign-in completes reliably with no 403.
2. Mode 0: ≥1,000 consecutive requests under 100 ms, zero 403.
3. Modes 2 & 3: ≥1,000 consecutive requests each under 500 ms, no 403 and no 429.
4. Logging proves sustained performance and steady data flow for at least two hours of parallel sessions.

Hand over the updated code, any new config files, and a brief README that explains how to spin up, test, and monitor the service. If you are comfortable diving deep into API quirks, concurrency limits, and micro-optimisation inside Node.js, I’m ready to work with you right away.