paper-from-zero

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

• Third-party content exposure detected (high risk: 0.80). The skill's workflow (SKILL.md step 2 "Actively search literature and build a compact literature map") and the Collaboration Hook #1 (call to Gemini literature-map after step 2) explicitly require ingesting candidate papers/keyword clusters from external/public literature, which the agent reads and uses to build the evidence matrix and routing/claim decisions—exposing it to untrusted third‑party content that can influence actions.