ai-second-brain

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

• Third-party content exposure detected (medium risk: 0.30). The required workflow ingests outsider-authored text from GitHub gists via curl and WebFetch, as well as from Gmail, iMessage, and NotebookLM sources connected to the second brain.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill fetches a setup instruction file from a GitHub Gist (gist.githubusercontent.com/karpathy/442a6bf555914893e9891c11519de94f/raw) during runtime execution, representing an unverified third-party content dependency under weak/unknown provenance.