getting-started

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

• Third-party content exposure detected (high risk: 0.90). FLAG: SKILL.md explicitly instructs the agent to fetch and analyse public web and social media content (e.g., the "Employer Footprint" step that "scan[s] Google, LinkedIn, Twitter/X", "Application Optimiser" research including Glassdoor reviews, and "Social Media Review" that "searches for your name and handles"), and that untrusted third‑party content is read and used to drive research briefs, CV rewrites, interview prep, and other decisions.