production-grade

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly requires the agent to "harvest every reachable surface" and "check current official docs" and to read connected MCPs and peer-skill catalogues (see M2 "Before acting, the agent harvests every reachable surface" and references/06 routing table with external GitHub/docs links), which means it will fetch and interpret open/public third‑party content that can influence decisions and tool use.