Powered by x402 Protocol

Secure Every
Agent Skill

Pre-install security auditing for AI agent skills. YARA malware detection, permission analysis, and trust attestation — all before you install.

286+
Skills Scanned
<2s
Avg Scan Time
$0.10
Starting Price
The Problem

The Risk of Unsigned Skills

"skill.md is an unsigned binary" — scanning 286 ClawdHub skills revealed credential stealers disguised as legitimate tools.

Hidden Malware

Credential stealers disguised as legitimate skills, silently reading your .aws/credentials and .ssh keys.

Data Exfiltration

Skills that POST your secrets to webhook.site or external servers without any visible indication.

No Pre-Install Checks

x402-secure handles payments, but nobody verifies the skill code BEFORE you install it.

Excessive Permissions

Skills requesting full system access when they only need to fetch weather data.

Suspicious Network Calls

Outbound connections to unknown IPs or domains that have nothing to do with the skill's purpose.

Obfuscated Code

Base64-encoded payloads and eval() calls hiding malicious intent in plain sight.

AI Agent Skills

What Powers Modern AI Agents

SKILL.md files teach AI agents how to use tools. They can execute commands, access files, and control browsers — that's why security auditing is essential.

File Access

Read and write files on your system

  • Create documents
  • Modify configs
  • Access data files

Shell Commands

Execute terminal commands directly

  • Run scripts
  • Install packages
  • Manage processes

Browser Automation

Control web browsers programmatically

  • Open URLs
  • Fill forms
  • Capture screenshots

Why Skills Need Auditing

Without security auditing, you're trusting unknown code with system access

Without Auditing

Credential Theft

Malicious skills can steal API keys and passwords

Data Exfiltration

Sensitive data can be sent to external servers

Malicious Commands

Hidden commands can damage your system

Excessive Permissions

Skills may request more access than needed

With x402guard

Permission Verification

Every permission request is analyzed and flagged

Network Inspection

All external calls are detected and reviewed

Malware Detection

YARA rules catch known malicious patterns

Safe Recommendations

Clear risk scores help you decide safely

The SKILL.md Format

A teaching document that tells AI how to use a specific tool

skill.md
---
name: nano-banana-pro
description: Generate images via Gemini
metadata:
  openclaw:
    requires:
      bins: ["uv"]
      env: ["GEMINI_API_KEY"]
---

# Instructions

This skill enables image generation using
Google's Gemini API. The agent can create
images based on text descriptions.

## Usage
Describe the image you want to generate...

Security-Relevant Fields

name

Unique identifier for the skill

Low
description

What the skill does

Low
bins

Required system binaries (executables)

High Risk
env

Required environment variables (API keys)

High Risk

Real AI Agent Capabilities

These powerful capabilities are why security auditing matters

“I wanted to automate tasks from Todoist and claw was able to create a skill for it on its own, all within a Telegram chat”

@iamsubhrajyotiSkill self-creation

“My OpenClaw realised it needed an API key... it opened my browser... opened Google Cloud Console... Configured OAuth and provisioned a new token”

@InfoxicadorAutonomous API setup
The Solution

x402guard fills the security gap

The missing Layer 2 of the agentic trust stack. While x402-secure handles payment security, x402guard verifies code security BEFORE installation.

YARA-based malware detection
Permission manifest extraction
Network call analysis
Credential access detection
Risk scoring (0-100)
Actionable recommendations

Agentic Trust Stack

Layer 4

Payment Security

x402-secure
Layer 3

Runtime Behavior

Trustline
Layer 2

Code Security

x402guard ★
Layer 1

Identity

ERC-8004

How It Works

1

Submit Skill

Provide a skill URL or paste the content directly

2

Deep Analysis

YARA scans, permission extraction, network analysis

3

Risk Assessment

Get a risk score (0-100) with detailed findings

4

Attestation

Receive a signed attestation for verified skills

Try It Now

Scan a Skill

Pay per audit with USDC on Base via x402. Connect your wallet, sign the payment, and get instant security results.

Connect Your Wallet

Connect a wallet with USDC on Base to pay for audits

Quick

Fast YARA scan

$0.10USDC
  • YARA malware detection
  • Risk score (0-100)
  • Risk level classification
  • Basic recommendation
Most Popular

Standard

Full analysis

$0.50USDC
  • All Quick features
  • Permission analysis
  • Network call detection
  • Detailed findings report

Deep

Complete audit

$1.00USDC
  • All Standard features
  • Behavioral sandbox
  • Signed attestation
  • Full audit trail

Enter the URL of the skill to scan

Or paste the skill content directly

For Developers

Simple Integration

One API call with x402 payment. Works with any agent framework, any language, any platform.

Terminal
# Quick scan ($0.10)
curl -X POST http://x402guard.xyz/audit/quick \
  -H "Content-Type: application/json" \
  -H "X-Payment: <x402-payment-token>" \
  -d '{"skill_url": "https://clawdhub.com/skills/weather"}'

# Standard scan ($0.50)
curl -X POST http://x402guard.xyz/audit/standard \
  -H "Content-Type: application/json" \
  -H "X-Payment: <x402-payment-token>" \
  -d '{"skill_url": "https://clawdhub.com/skills/weather"}'

# Deep scan ($1.00)
curl -X POST http://x402guard.xyz/audit/deep \
  -H "Content-Type: application/json" \
  -H "X-Payment: <x402-payment-token>" \
  -d '{"skill_url": "https://clawdhub.com/skills/weather"}'

Response

JSON Response
{
  "risk_score": 12,
  "risk_level": "LOW",
  "recommendation": "SAFE",
  "findings": {
    "malware": [],
    "permissions": ["network:read"],
    "network": ["api.weather.com"]
  },
  "audit_id": "aud_abc123",
  "timestamp": "2026-01-31T10:30:00Z",
  "tier": "standard",
  "attestation": "0x..."
}

x402 Integration

Learn how to integrate x402 payments into your agent for seamless audits.

x402 Protocol

Enterprise Solutions

Need custom integration or volume pricing? DM @goheesheng on X, founder of x402guard.

DM on X
Questions?

Frequently Asked Questions

Everything you need to know about x402guard

Still have questions? We're here to help.

Contact Support