tako

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

• Third-party content exposure detected (high risk: 0.85). The skill's Knowledge Search explicitly supports a "web" source_index and returns webpage_url/embed_url and extracted descriptions/insights (see SKILL.md examples and the "source_indexes: ['tako', 'web']" / "Empty knowledge_cards" tip), so it ingests open/public third-party web content that could contain untrusted, user-generated instructions.