validation-expert

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

• Third-party content exposure detected (high risk: 0.80). The skill instructs launching research subagents in Step 2 ("codex-deep-research") and Step 3 ("gemini-research-analyst") to "research" industry best practices and external tools—tasks that will ingest and interpret open/public web content (third‑party sites, docs, and company examples), exposing the agent to untrusted user-generated/third‑party material.