Nooks: Places worth revisiting

Personal place intelligence — save and find cafes, coworking spaces, libraries, restaurants, and other spots worth revisiting. Stores one markdown file per p...

This skill appears coherent and low-risk for its stated purpose, but check a few things before installing: 1) the agent will read and modify files in mind/nooks/ — don't keep secrets there; 2) if you enable Maps automation, store your Google Places API key only in mind/nooks/nooksconfig.yml and restrict the key to the Places API and appropriate referrers/IPs; 3) the README suggests fetching SKILL.md via curl from a GitHub raw URL — verify you trust that repository before running the command; 4) the skill recommends adding a heartbeat/cron (every 30 minutes) which increases how often it will access your files — only enable if you want periodic checks; 5) if you prefer to avoid autonomous background activity, disable autonomous invocation for this skill in the agent settings. If you want more assurance, provide the upstream GitHub repo for verification or inspect the SKILL.md/README in your environment before installing.

Type: OpenClaw Skill Name: nooks Version: 1.0.1 The skill includes a self-update mechanism in SKILL.md that instructs the agent to fetch and replace its own logic from a remote GitHub URL (raw.githubusercontent.com), which is a vector for Remote Code Execution (RCE). Additionally, it suggests establishing persistence by creating a cron job or modifying HEARTBEAT.md for automated execution every 30 minutes. While these features are aligned with the stated purpose of maintaining a 'place intelligence' database, they represent significant security risks common in supply-chain attacks.