conducting-api-security-testing

SUSPICIOUS: the skill is internally consistent as an API pentesting guide and shows no clear credential exfiltration or malicious installer behavior, but it grants an AI agent offensive security capabilities against external APIs. The scanner’s command-injection hit is a false positive from markdown, yet the overall skill remains high risk because it operationalizes penetration testing techniques.