SafeAgent-BeihangSafeAgent-BeihangOfficialยท6 Agent Skills Included

clawguard

Runtime security auditing, threat detection, and injection defense

Audits skills before installation, monitors runtime behavior, and blocks prompt injection attacks. Detects intent drift, reverse shells, credential theft, and data exfiltration without manual log review. Freezes risky sessions, preserves evidence, and auto-learns new blocking rules from detected threats.
npx skills add SafeAgent-Beihang/clawguard --all -g -y

All Skills in This Repository (6)

Pure Emerald Level Indicators

Frequently Asked Questions

FAQPage Schema
How to install ClawGuard?โ–ผ

Run `npx skills add SafeAgent-Beihang/clawguard --all -g -y` in your terminal to install all five security modules globally.

How to check if a skill is safe before installing?โ–ผ

Ask your agent to audit the skill, and the Auditor module scans its code, dependencies, and stated purpose to flag hidden risks like SSH key access or data exfiltration.

Can ClawGuard block prompt injection attacks?โ–ผ

Yes. The Shield module detects jailbreak attempts, role hijacking, zero-width characters, and encoded commands in user inputs, then blocks or sanitizes them.

Does ClawGuard monitor agents in real time?โ–ผ

Yes. The Detect and Guardian modules watch commands, file access, and network traffic live, and can freeze a session instantly when a critical threat appears.

Can I use ClawGuard without security expertise?โ–ผ

Yes. Just describe what you want in plain English, such as 'check my config security', and the agent follows the built-in guides to run checks and suggest fixes.

Related Repositories in Software Engineering

View All in Software Engineeringโ†’