United States

Plan United States trips with region-specific routing, verified entry rules, transport logistics, and practical tourist safety.

This skill appears coherent and self-contained, but review these practical points before installing: 1) The skill writes and reads a local memory file at ~/united-states/memory.md — avoid storing highly sensitive data (full passport numbers, raw copies of visas, full payment credentials) in that file. 2) The SKILL.md asserts it will not access files outside ~/united-states/ or make network calls, but that is an instruction; confirm your agent/runtime enforces those policies if you require strict privacy. 3) The agent will 'read it silently' when the memory file exists — understand that previously stored preferences and statuses will be used automatically unless you clear or inspect the file. 4) Related-skills installs (clawhub install ...) are opt-in; do not install additional skills unless you trust their source. If you want extra assurance, open and inspect ~/united-states/memory.md after initialization and remove any sensitive tokens before proceeding.

Type: OpenClaw Skill Name: united-states Version: 1.0.0 The skill is designed for travel planning and primarily uses local file system operations to manage trip memory within `~/united-states/`. It contains a direct shell command (`mkdir -p ~/united-states`) in `setup.md` to create its workspace. While this specific command is benign and necessary for the skill's stated purpose, the capability for the AI agent to execute shell commands from markdown instructions represents a potential attack surface (e.g., RCE risk) if not carefully constrained or if user input could influence the command. The `SKILL.md` explicitly states the skill does NOT make network requests, and there is no evidence of data exfiltration, backdoors, or malicious prompt injection.