content-creation

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.80). These links mix legitimate platform APIs with unvetted third‑party GitHub repos, MCP endpoints and a raw install script (raw.githubusercontent.com/.../install.sh) plus direct media/download endpoints — meaning downloading or piping installers from them without verifying the source/signatures is risky.

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

• Third-party content exposure detected (high risk: 0.78). Outsider free text can enter the LLM context when the user provides a URL for “REFERENCE EXTRACTION,” since the skill uses curl/agent-browser to fetch and extract tweet/post text (e.g., curl https://api.fxtwitter.com/status/$TWEET_ID and agent-browser get text body) and then sends that extracted text to Gemini/Claude for analysis and style-brief generation.