SafeAI-Lab-X

SafeAI-Lab-X / ClawKeeper

Public

ClawKeeper: Comprehensive Safety Protection for OpenClaw Agents Through Skills, Plugins, and Watchers (aka The Norton for OpenClaw)

agents openclaw plugins safety security
21
1
89% credibility
Found Mar 26, 2026 at 22 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

ClawKeeper is a multi-layered security framework for OpenClaw AI agents using skills, plugins, and watchers to prevent threats like prompt injection, leaks, and drifts.

How It Works

1
🔍 Discover safety for your AI helper

You find ClawKeeper while looking for ways to keep your OpenClaw AI agent secure from risks like bad instructions or leaks.

2
📦 Add the safety tools

Run a simple installer to add protections like checks, rules, and watchers that guard your setup.

3
🛡️ Check for weak spots

Run a quick scan to see your security score and spot any easy fixes.

4
✨ Strengthen automatically

One-click applies safe fixes, backs up your setup, and boosts your protection without hassle.

5
📱 Add smart rules and monitoring

Load easy rules for your platforms and start ongoing watches for drifts or threats.

✅ Your AI is now safely guarded

Enjoy your OpenClaw agent with strong defenses, clear reports, and peace of mind.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 22 to 21 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is ClawKeeper?

ClawKeeper delivers comprehensive safety protection for OpenClaw agents through skills, plugins, and watchers—aka the Norton for OpenClaw. Built in TypeScript, it layers instruction-level safeguards via Markdown skills, runtime enforcement with plugins for auditing and hardening, and external watchers for real-time monitoring and intervention. Users get CLI tools like `npx openclaw clawkeeper audit`, `harden`, `monitor`, and `scan-skill` to scan configs, block threats, log events, and rollback changes without touching agent code.

Why is it gaining traction?

It stands out by combining proactive threat blocking (prompt injection, credential leaks, risky tools) with easy automation like one-click hardening and nightly audits, outperforming basic OpenClaw security repos in benchmarks. Developers notice seamless cross-platform support (Windows, Linux, macOS) and Feishu integration, plus detailed logs queryable by tool or type—no more manual hunting for anomalies.

Who should use this?

OpenClaw users deploying agents in production, especially those handling sensitive data or third-party skills. Ideal for AI ops engineers securing gateways, devs scanning plugins for supply chain risks, or teams needing behavioral profiling to catch intent drift in multi-turn sessions.

Verdict

Grab it if you're running OpenClaw seriously—solid docs and MIT license make it production-ready despite 21 stars and a 0.90% credibility score signaling early maturity. Pair with the bundled skills for quick wins, but test watchers in staging first.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.