conducting-api-security-testing

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

• Third-party content exposure detected (high risk: 0.80). The agent's required runtime workflow directly issues HTTP requests to an arbitrary target (--base-url) and programmatically ingests and interprets responses (see scripts/agent.py functions like test_graphql_introspection, test_excessive_data_exposure, test_bola and the SKILL.md / references showing importing a target API base URL), so it clearly consumes untrusted third‑party API content that can influence its test decisions.