Openclaw User Profiler

Two things: (1) Build a user.md through conversation so your OpenClaw lobster knows who it's working with. (2) Recommend Claude Code Skills based on the user...

This skill appears to do what it says: it will ask you for a directory, read an existing user.md if present, let you update it, and write a user.md file. Before installing/using it, consider: (1) it needs permission to read and write files in directories you point it to — don't point it at sensitive locations. (2) it will scan your Claude skills directory to see what's installed; that requires filesystem access but not credentials. (3) recommendations may include many third-party/community Skills that are installed using commands like `npx skills add <package>` — installing those external packages is a separate security decision (they may run code). (4) the skill explicitly says it will refuse to store sensitive info, but avoid pasting secrets or private IDs into the profiling conversation. If you want extra assurance, review the linked GitHub repo (homepage) to confirm authorship and examine the catalog of recommended packages before following any install suggestions.

Type: OpenClaw Skill Name: openclaw-user-profiler Version: 2.3.1 The openclaw-user-profiler skill is designed to create a local user profile (user.md) and recommend relevant AgentSkills based on the user's professional role. The code and instructions in SKILL.md and the reference files focus entirely on conversational data gathering and matching roles to a curated catalog of third-party skills. Notably, the documentation in references/user-md-template.md includes explicit security warnings advising the agent and user never to store sensitive information like passwords or API keys. There is no evidence of data exfiltration, obfuscation, or malicious prompt injection intended to bypass safety protocols.