prompt-engineer

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

• Third-party content exposure detected (low risk: 0.10). The workflow’s “Test and evaluate” stage consumes outsider-authored free text if the test suite inputs are populated from user-submitted sources (e.g., the evaluation runner formats case["input"]/test_case["question"] directly into prompts for LLM completion and judge scoring), but no such external ingestion path is specified in the provided skill docs.