clean-code

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

• Third-party content exposure detected (high risk: 0.80). The skill requires running verification scripts that accept arbitrary URLs (e.g., performance-optimizer's lighthouse_audit.py <url> and test-engineer's playwright_runner.py <url>) and mandates reading/parsing their output ("Run the script and capture ALL output" → "READ → SUMMARIZE → ASK"), which means the agent will ingest and act on untrusted third-party web content.