bugzilla

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

• Third-party content exposure detected (high risk: 0.80). This skill queries and prints user-generated content from the public Bugzilla site (see scripts/bz.py calls to GET endpoints like bug/{id} and bug/{id}/comment and the search command that returns bug summaries/comments), so the agent will read/display untrusted third-party bug reports and comments that could contain injected instructions.