Enterprise-grade 24/7 desktop AI assistant with system-level fault tolerance, auto-restart, performance monitoring, intelligent voice, and resource optimizat...

This skill is instruction-only and tells you to clone and run an external GitHub project (npm install, pip install edge-tts, npm start) and edit local files to integrate with OpenClaw on localhost:18788. The registry metadata should have listed required binaries and the fetching/installation step but did not — that's the core inconsistency. Before installing: 1) review the GitHub repository and source code yourself (or have someone you trust do it); 2) inspect any npm/pip dependencies for malicious packages; 3) avoid running installs as administrator and consider running inside a VM or sandbox first; 4) verify the app's behavior around automatic restarts/logging and that it does not open unexpected network connections; 5) confirm the OpenClaw integration (port 18788) is intended for your environment. The changelog/metadata contain odd small details (very short development time) that suggest you should be extra cautious. If you want a safer allowlist, ask the author to provide a registry install spec and explicitly declare required binaries and permissions.

Type: OpenClaw Skill Name: claw-desktop-pet Version: 1.3.0 The skill bundle instructs the OpenClaw agent to perform a `git clone` from a public GitHub repository, followed by `npm install` and `pip install` to fetch and execute external code and dependencies. This introduces a significant supply chain risk, as the agent is directed to set up a full development environment and run arbitrary code from an external source. Additionally, the `QUICKSTART.md` file provides an example demonstrating the agent's ability to execute local shell commands (`await exec(...)`), highlighting broad system access capabilities required by the skill. While these actions are plausibly needed for the stated purpose of a desktop pet with system monitoring, they represent high-risk capabilities without clear evidence of intentional malicious behavior.