0xtresser

0xtresser / skill-guardian

Public

86,948 agent skills exist. Under 3% are worth installing. Skill Guardian intercepts skill installations to scan for security threats, detect duplicates, and assess necessity — so only high-value skills make it into your agent's context window.

agent agent-skills ai openclaw
17
0
89% credibility
Found Mar 13, 2026 at 17 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
AI Summary

Skill Guardian is a protective skill for AI agents that intercepts and vets new skill installations for security risks, duplicates, and quality before allowing them.

How It Works

1
😩 Feeling overwhelmed by bad skills

You notice your AI helper keeps suggesting or installing skills that are empty, duplicate, or even risky, making it slower and less reliable.

2
🔍 Discover Skill Guardian

You learn about this helpful guardian that protects your AI helper by checking skills before adding them.

3
🚀 Ask your AI to add the guardian

Simply tell your AI helper to install Skill Guardian, and it sets itself up quickly with no hassle.

4
🛡️ Try installing a new skill

Next time you want a new skill, your guardian steps in to scan it for dangers and check if it's really needed.

5
Review the check results
✅
Approve and add

You say yes, it safely installs the skill, and your helper gets better without the junk.

❌
Block it

You decide no, it stops the install to keep things clean and secure.

🏆 Enjoy a smarter, safer helper

Now your AI only has the best skills, runs faster, thinks clearer, and stays protected from bad ones.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 17 to 17 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 skill-guardian?

Skill Guardian is a lightweight agent skill that intercepts new skill installations in your AI agent's ecosystem, where 86,948 agent skills exist but under 3% are high-value. It scans for security threats using tools like snyk-agent-scan and skills.sh audits, detects duplicates, and assesses necessity to avoid bloating your agent's context window. Implemented as a pure SKILL.md file with zero dependencies, it forces a report-and-confirm step before any install.

Why is it gaining traction?

Unlike post-install scanners like snyk-agent-scan or passive audit lists from skills.sh, Skill Guardian acts as a pre-install gate with necessity checks—spotting if an MCP tool already covers it or quality signals are poor. Developers notice fewer hallucinations, less context bloat, and cleaner installs right away via simple commands like `npx skills add 0xtresser/skill-guardian`. The cleanup mode for auditing existing skills seals the deal for maintaining lean agent setups.

Who should use this?

AI agent users on platforms like Claude Code, Cursor, or Windsurf who frequently add skills via `npx skills add` and want to dodge the 36.8% with vulnerabilities or 28% duplicates. Ideal for devs building agent workflows tired of sifting through junk in the skills dashboard, ensuring only essential tools load.

Verdict

Grab it if you're deep in agent skills—solid concept for safer, slimmer contexts despite 17 stars and 0.9% credibility signaling early days with minimal code beyond docs. Test in a sandbox first; maturity lags, but the MIT license and one-file install make it low-risk to evaluate.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.