ramp

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

• Third-party content exposure detected (high risk: 0.85). The skill’s runtime context ingestion path includes reading ~/.claude/plugins/config/claude-for-legal/legal-clinic/CLAUDE.md and the “ingested handbook” uploaded by the professor; if those files contain professor-authored text, that is outsider-authored free text (professor is not the operating user) that the LLM reads into context.