report-render

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill requires copying "content already present in the conversation" verbatim into a generated HTML report (no rewriting), so any API keys/passwords or other secrets present in the conversation would be reproduced exactly in the output — creating a high exfiltration risk.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The scaffold embeds and imports remote scripts that will be fetched and executed at runtime in the generated report (https://cdn.tailwindcss.com and https://cdn.jsdelivr.net/npm/mermaid@11/dist/mermaid.esm.min.mjs), so the skill relies on external code that runs in the user's environment.