vibes

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt instructs the agent to ask the user for an OpenRouter API key and then embed that key verbatim into deployment CLI flags (e.g., --ai-key "sk-..."), which requires the LLM to handle and output secrets directly, creating exfiltration risk.

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

• Third-party content exposure detected (high risk: 0.90). The skill's Step 0.5 "Check for Design Reference" explicitly instructs the agent to read user-provided reference images or theme.html files via URL (WebFetch) and analyze that arbitrary third‑party content to guide design reasoning and code generation, which can materially influence subsequent actions.