nia-docs

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

• Third-party content exposure detected (high risk: 0.80). The skill performs searches across public package registries and indexed sources (explicitly "3000+ packages (npm, PyPI, Crates, Go)" and other indexed documentation) and therefore ingests untrusted, user-generated third-party documentation and code that the agent will read and interpret.