clean-code

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill mandates capturing "ALL output" from verification scripts and summarizing it verbatim to the user (errors/warnings/passes) without any redaction guidance, which could force the LLM to include secrets or tokens present in script/log output.

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

• Third-party content exposure detected (high risk: 0.80). The skill's verification scripts include commands that accept arbitrary inputs (notably lighthouse_audit.py <url> and playwright_runner.py <url> in the Agent→Script Mapping) and the workflow mandates running those scripts and parsing their output, which means the agent will fetch and interpret content from open/public URLs (untrusted third‑party sources).