web-cache-poisoning
Warn
Audited by Socket on Sep 20, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
This skill is internally consistent but high-risk because its stated purpose is offensive web exploitation. External tool references are same-org/official PortSwigger resources, and there is no suspicious installer, credential harvesting, or hidden data exfiltration. The main concern is that it equips an AI agent to perform active cache-poisoning/deception attacks with real impact on third-party targets.
Confidence: 92%Severity: 78%
Audit Metadata