chatgpt-api

SUSPICIOUS: the skill is internally consistent with its stated reverse-engineering purpose, but that purpose itself relies on private ChatGPT web endpoints, hidden controls, and reuse of an authenticated browser session instead of official APIs. Data flows stay on official OpenAI domains, so this is not strong evidence of malware or credential harvesting, but it carries medium security risk due to session reuse, request interception guidance, and private-backend replay.