nix-init

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

• Third-party content exposure detected (medium risk: 0.30). evals/evals.json provides outsider-authored prompts that require nix-init to generate Nix from user-supplied repository URLs, and the workflow must consume that free-text prompt at runtime (indirect prompt-injection risk).

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

• Potentially malicious external URL detected (high risk: 0.90). The skill requires running nix-init with user-provided repository URLs (e.g., 'https://github.com/owner/repo' and https://github.com/ynqa/logu) which are fetched at runtime and whose contents directly determine the generated Nix derivation and can be built/executed, so these external URLs are runtime dependencies controlling the agent's output.