nixos

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.90). These URLs include a raw install script served from an untrusted GitHub user (curl | bash) and a URL shortener, both high-risk patterns for distributing malware or running arbitrary code.

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

• Third-party content exposure detected (high risk: 0.70). The required workflow in SKILL.md can trigger “Live Fetching” of documentation from raw GitHub URLs at runtime, which ingests outsider-authored free-form Markdown/text into the agent’s LLM context.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs the agent to perform live fetching of raw GitHub content at runtime (e.g. https://raw.githubusercontent.com/NixOS/nix.dev/master/source/.md) to source documentation that directly influences the agent's answers, meaning external content fetched at runtime controls agent instructions.