My in-house PHP proxy that pulls data from Semrush has suddenly stopped returning results. The script still runs, but every request comes back empty, so the issue is squarely in the data-fetching layer. I recently modified parts of the code—mainly the cURL header block and a small caching function—so I’m confident the breakage lives there rather than in Semrush itself. Here’s what I need from you: – Trace and identify the exact point where the script loses the payload (request formatting, response parsing, or proxy hand-off). – Patch the code so keyword, backlink, and audit calls all return complete JSON arrays exactly as Semrush delivers them. – Leave comments in the updated sections so future tweaks are simple. – Provide a short changelog outlining what was fixed and why. Acceptance is straightforward: I’ll run the same test queries that currently come back empty; when they return full datasets with accurate metrics, we’re done. You’re welcome to use any debugging tools you like—cURL, Postman, xDebug—as long as the final solution is clean PHP compatible with 8.1. If you’ve wrestled with Semrush endpoints, token limits, or proxy header quirks before, this will be quick work. Looking forward to your insight and a solid fix.