codexpro-chatgpt-local-agent

SUSPICIOUS: the skill is largely coherent with its stated purpose, but that purpose is inherently high-impact: it exposes a local repo to ChatGPT with write access, optional shell execution, public tunneling, and automated handoff execution. Install sources are partly trustworthy (official npm, documented Cloudflare tooling with checksums), yet publisher identity is mismatched and the runtime footprint is broader than a typical instructional skill. No clear credential theft or covert exfiltration is shown, so this is not confirmed malware, but it is a meaningful security risk.