load-testing

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The prompt asks the agent to collect M2M client_id/client_secret from the user and shows examples embedding them directly in CLI flags (and as direct exports), which would require the LLM to handle and potentially emit secret values verbatim in commands—an insecure pattern despite also mentioning env-vars.

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

• Third-party content exposure detected (high risk: 0.80). The skill's load-test scripts (locustfile.py and run_load_test.py) explicitly accept arbitrary --app-url endpoints and parse SSE streaming JSON responses from those apps as part of the required workflow (see SKILL.md: "locustfile.py — ... Sends POST /invocations ... Parses SSE stream (data: {json} lines)" and the run CLI that tests each app URL), so untrusted third-party app responses could influence test logic and metrics and enable indirect prompt injection.