resume-craft

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

• Third-party content exposure detected (high risk: 0.85). Phase 1/2/3/5 require the agent to ingest the user-provided JD text and resume text (which are outsider-authored free-form inputs from the operating user’s perspective), and then the agent feeds that readable JD/resume content into its LLM context for keyword extraction, matching, and STAR rewriting.