performing-api-fuzzing-with-restler

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt embeds plaintext credentials in the example authentication module and prints an Authorization header (causing tokens/passwords to be output or hard-coded into generated code/commands), which requires handling secrets verbatim and is high-risk.

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

• Third-party content exposure detected (high risk: 0.80). The skill's workflow and code (SKILL.md Step 2/3 and scripts/agent.py) explicitly require ingesting a provided OpenAPI specification and interacting with an arbitrary target API (the AuthenticationProvider's requests.post and RESTler runs via run_fuzz_mode), so untrusted, user-supplied API responses and specs are read and used to drive stateful fuzzing and subsequent tool actions.