clawhub

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

• Third-party content exposure detected (high risk: 1.00). This skill explicitly fetches and installs skills from the public ClawHub registry (see SKILL.md's "npx ... clawhub search" and "npx ... clawhub install" which place skills into ~/.nanobot/workspace/skills), meaning the agent will ingest and execute untrusted third‑party content that can change its behavior.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill instructs running npx --yes clawhub@latest (which fetches and runs remote package code from the npm registry at runtime), meaning external code is executed and can directly control agent behavior when installing/loading skills.