perplexity-search

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill includes examples that place API keys directly into command-line arguments and explicit export lines (e.g., --api-key sk-or-v1-your-key-here and export OPENROUTER_API_KEY='sk-or-v1-your-key-here'), which require the LLM or user to handle and/or output secret values verbatim and expose them in shell history/process lists.

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

• Third-party content exposure detected (high risk: 0.85). Yes—at runtime the script sends the user-provided query as the LLM prompt, and Perplexity/OpenRouter performs web-grounded retrieval whose resulting web text is incorporated into the model’s generated context/answer (outsider-authored public web content).