The Agent Skills Directory

[SAFE]: No security issues detected. The skill contains documentation, configuration advice, and best practices for the OpenClaw framework.- [SAFE]: The installation command uses npx to add skills from the author's own repository (ManiClones/agent-skills), which is a legitimate vendor resource.- [SAFE]: The skill mentions local file paths like ~/.openclaw/agents/ and files like PLAN.md for workflow management, which are standard for the described application and do not involve unauthorized data access.- [SAFE]: References to external services like Claude, Gemini, and Ollama are for legitimate AI model providers as part of the suggested architecture.

10-things-i-wish-i-knew-before-using-openclaw