Back to plugin

Security audit

Tinker Learned Intuition

Security checks for vulnerabilities and agentic risk

Overview

This plugin is a disclosed, opt-in safety gate that observes tool calls by default and only actively blocks or mutates prompts when the user enables those settings.

Installers should understand that the default mode observes and logs rather than blocks. Only enable hookEnforcement if you intentionally want a machine-wide pre-tool safety gate, and only enable personalityNudge if you are comfortable with another plugin later injecting behavior-steering text into prompts. Review the local ~/.openclaw logs if tool metadata or command text may be sensitive.

SkillSpector was not run because this plugin release contains no bundled skills.

Static analysis

No suspicious patterns detected.