performing-web-cache-poisoning-attack

The skill is internally consistent with its stated purpose, but that purpose is an offensive security capability for an AI agent. It enables active exploitation of targets, including poisoning shared caches and testing authenticated content leakage, creating high real-world abuse risk even though the static data-exfiltration findings themselves are only documentation-context examples.