nlm-skill

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The prompt instructs users to extract and paste authentication cookies (mcp__notebooklm-mcp__save_auth_tokens(cookies="<cookie_header>")), which would require the agent to accept and include sensitive cookie/header values verbatim in a tool call or output, creating a high exfiltration risk.

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

• Third-party content exposure detected (high risk: 1.00). The SKILL.md workflows explicitly ingest open web and user-generated content (e.g., "nlm source add --url 'https://...'", YouTube URLs, and "nlm research start 'query' --notebook-id ") and then have the agent read/describe/query those sources and use them to generate artifacts (studio/report/audio) and drive follow-up actions, so untrusted third‑party content can materially influence tool behavior and pose an indirect prompt-injection risk.