parallel-web

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.80). The URL https://parallel.ai/install.sh is suspicious because it's a direct shell installer referenced in a curl|bash install command (high-risk to run unverified remote scripts); the other URLs are academic/documentation sites and are low-risk.

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s runtime workflow for Web Search/Web Extract/Deep Research (e.g., parallel-cli search, parallel-cli extract, parallel-cli research run), triggered by the user’s natural-language request, causes the LLM to ingest outsider-authored free text from search results and fetched webpages/contents without selecting a specific trusted item first, creating indirect prompt-injection exposure.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill's setup includes a runtime command that fetches and executes remote code via curl -fsSL https://parallel.ai/install.sh | bash, which downloads and runs a remote install script.