reverse-engineer

The skill describes detailed, practical steps to extract session tokens and interact with undocumented Perplexity AI endpoints and to integrate findings into an SDK. While the technical steps are reasonable for authorized reverse-engineering and internal SDK development, the explicit advice to extract and reuse pplx.session-id cookies and to employ anti-detection/fingerprinting techniques creates a substantial abuse risk. There is no embedded malware or obfuscated code in the artifact, but its operational guidance could facilitate credential theft, unauthorized access, and covert scraping if used without strict authorization and controls. Recommend restricting use to authorized research with documented consent, using test/dev accounts, secure secret handling, and removing or gating anti-detection instructions.