nemoclaw-user-configure-security

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

• Third-party content exposure detected (high risk: 0.80). The skill's documentation (references/openclaw-controls.md and SKILL.md) explicitly describes presets and network endpoints (e.g., github, pypi, npm, discord CDN, huggingface, brave search) and an operator-approval/web-fetch flow where web fetch/search results are "auto-wrapped as untrusted external content," showing the agent can fetch and ingest public third-party pages/search results that may influence tool calls and decisions.