Is icemastert/agent-builder-1-0-0 safe?

https://github.com/openclaw/skills/tree/main/skills/icemastert/agent-builder-1-0-0

98
SAFE

This is a legitimate OpenClaw agent builder skill that provides documentation and templates for creating AI agent configurations. The skill contains only markdown files with no executable code, shows no signs of malicious behavior, and includes security-conscious practices in its templates.

Category Scores

Prompt Injection 95/100 · 30%
Data Exfiltration 100/100 · 25%
Code Execution 100/100 · 20%
Clone Behavior 95/100 · 10%
Canary Integrity 100/100 · 10%
Behavioral Reasoning 95/100 · 5%

Findings (3)

LOW References to external file reads -5 ▶

The skill instructs the agent to read reference files, but these are local files within the skill directory and pose minimal risk

LOW Expected GitHub network connections -5 ▶

Network connections to GitHub detected during git clone operation, which is expected behavior for skill installation

INFO Security-conscious template practices -5 ▶

The skill templates include good security practices like requiring permission before destructive actions