revops

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). Prospects can submit free-text demo or contact form fields (company, job title, message) that the enrichment and routing workflow ingests automatically at form submit, without the agent selecting a specific item, so outsider text reaches the agent through the required inbound path.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (medium risk: 0.30). The skill's own operational path directs the agent to read named instruction files from linked Quegenx/agent-skills GitHub URLs when the installed folder is unavailable, and Quegenx is an attributable third-party skill collection with no strong first-party provenance or enforced digest, so the dependency is common/reputable unofficial and rates medium.