Back to skill
Skillv1.2.1
VirusTotal security
Agent Hush · External malware reputation and Code Insight signals for this exact artifact hash.
Scanner verdict
BenignApr 30, 2026, 6:19 AM
- Hash
- c3328cc57d2f2d057a49c1b2d903cae23d96f6c825dc173bb9da1340e6fbc268
- Source
- palm
- Verdict
- benign
- Code Insight
- Type: OpenClaw Skill Name: agent-hush Version: 1.2.1 Agent Hush is a privacy-focused security utility designed to detect and redact sensitive information (API keys, tokens, PII, and infrastructure details) before it is leaked through outbound actions like git pushes or skill publishing. The core logic in `scripts/sanitize.py` implements a comprehensive scanner using over 200 regex rules derived from the Gitleaks project and AI-specific patterns, operating entirely locally with no external network dependencies. The instructions in `SKILL.md` guide the AI agent to proactively and silently intercept high-risk actions to perform these checks, which is consistent with the tool's stated purpose as a privacy guardian.
- External report
- View on VirusTotal