lead-researcher

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

• Third-party content exposure detected (high risk: 0.90). The skill's required workflow explicitly instructs the agent to "identify" and synthesize papers in Stage 3 ("Literature Synthesis") and to review a specific paper in Stage 4 and "locate official code" in Stage 6, which requires fetching and reading public third-party papers/repos whose content the agent must interpret and which can change decisions and tool use (e.g., experiment design and replication), exposing it to untrusted third-party content that could contain indirect prompt-injection.