ai-forge-judge

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

• Third-party content exposure detected (low risk: 0.10). The required runtime workflow only ingests a target prompt supplied to the evaluator (e.g., this SKILL.md file) and bundled first-party reference files from the local references/ directory (optionally WebFetch a spec URL), with no runtime path that reads outsider-authored feeds/queues/tickets or arbitrary user content beyond the target being evaluated.