skills.rest
yuanyuekejiJN
yuanyuekejiJNCOMMUNITY
Explore all skills by this author →
GitHub

clawdefender

Detect and mitigate injection, exfiltration, and path traversal threats in AI agent workflows.

18|2|Updated Mar 14, 2026
Search Tags:#security#url validation#input sanitization#prompt injection#malware scanning#ssrf protection#workspace audit
One-click install
npx skills add https://github.com/yuanyuekejiJN/AivoClaw --skill clawdefender-yuanyuekejijn
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: clawdefender
Source: https://github.com/yuanyuekejiJN/AivoClaw/tree/main/extraSkills/clawdefender-1
Command: npx skills add https://github.com/yuanyuekejiJN/AivoClaw --skill clawdefender-yuanyuekejijn

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires bash, grep, sed, jq, npx, and includes scripts (resource) components.

What problem does it solve?

This Skill helps protect AI agents from malicious or unsafe content by detecting prompt injection, command injection, SSRF/data exfiltration attempts, credential theft patterns, and path traversal indicators before the agent processes untrusted data.

Core Features & Use Cases

  • Security audit for skills: Scans installed skills and workspace scripts to flag potentially dangerous patterns with severity levels.
  • Universal input sanitization: Checks external text payloads (including email/calendar/API responses) for prompt-injection indicators and can wrap/label flagged content.
  • URL validation to prevent SSRF/exfiltration: Detects unsafe hostnames/IP ranges and known exfiltration endpoints before fetching.
  • Automation-friendly checks: Supports non-interactive validation modes such as strict failure (exit code 1) and report-only output.

Quick Start

Run a security audit by instructing your agent: execute the ClawDefender full audit command to scan all installed skills and scripts for critical and high-risk findings.

Frequently Asked Questions about clawdefender

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I detect prompt injection in untrusted text before my AI agent processes it?▼

Prompt injection detection scans external text payloads like email or API responses for malicious indicators, applying severity scoring and optional strict blocking to prevent unsafe content from reaching your agent.

What is the best way to scan installed skills and scripts for command injection and credential theft patterns?▼

Security auditing scans workspace files and installed scripts to flag dangerous patterns like command injection and credential theft, assigning severity levels to identify critical and high-risk findings.

How do I prevent SSRF and data exfiltration when fetching URLs in an AI workflow?▼

URL validation prevents SSRF and data exfiltration by detecting unsafe hostnames, IP ranges, and known exfiltration endpoints before your agent fetches external resources.

Can I run automated security checks on AI agent inputs without manual review?▼

Automated checks support non-interactive validation modes including strict failure with exit code 1 and report-only output, enabling seamless integration into CI pipelines and agent workflows.

Do I need bash and jq installed to run workspace security scanning for path traversal threats?▼

Yes, workspace security scanning requires bash, grep, sed, jq, and npx to execute pattern-based threat detection across installed skill files and identify path traversal indicators.

How does severity scoring work when sanitizing external payloads for prompt injection indicators?▼

Severity scoring evaluates detected prompt injection patterns in external payloads, categorizing threats by risk level to determine whether to wrap, label, or strictly block the flagged content.

Related Skills

Explore More Legal & Compliance Skills→
vercel-labsvercel-labs
609

text

Analyze text files with bash utilities for statistics, pattern searches, and word frequencies.

Official
Basic
AcczdyAcczdy
2

testing-for-xml-injection-vulnerabilities

Test XML injection vulnerabilities via HTTP requests and generate JSON findings reports.

Community
Advanced
archview-aiarchview-ai

openwebf-security-xss-sanitization

Sanitize untrusted HTML input to mitigate XSS risks in WebF applications.

Official
Advanced
EngineerWithAIEngineerWithAI
4

sast-configuration

Configure Semgrep, SonarQube, and CodeQL for automated vulnerability detection.

Official
Intermediate