flow-nexus-swarm

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly spawns "researcher" agents with a "web_search" capability and includes a "Research & Analysis" pattern and GitHub-triggered workflows that describe distributed information gathering, indicating the agents will fetch and interpret public third-party web content as part of their workflow and thus could be influenced by external (untrusted) sources.