Skill Security Audit Service
Five-tier Audit · Six-dimension Analysis · LLM Review
Audit a Skill
Enter a ClawHub or GitHub skill URL to check for security issues
-
Total Audits
-
Safe
-
Suspicious
-
Malicious
Audited Skills
0
requests today
Skill
Source
Result
Risk Score
Five-tier Detection
Core Rules → Dynamic Rules → LLM Analysis → Sandbox → LLM Review
Sandbox Execution
Firecracker microVM isolation for safe dynamic analysis
Rule Evolution
Continuous learning of new attack patterns with automatic rule promotion
SAFE
0.2
Risk
Detected Threats
LLM Semantic Analysis
Inferred Intent
-
Risk Factors
Recommendation
-
Sandbox Execution
Security Behavior
Performance Metrics
Audit Statistics
1,247
Total Audits
892
Safe
298
Suspicious
57
Malicious
7-Day Audit Trend
Threat Distribution
Recent Audits
Skill
Source
Result
Time
API Reference
Base URL: https://api.clawsec.cc
Endpoints
POST
/api/v1/audit
—
Five-tier six-dimension security audit
POST
/api/v1/api-keys/register
—
Agent registration (GitHub Device Flow)
POST
/api/v1/api-keys/verify
—
Poll for API key after user authorizes
GET
/api/v1/stats
—
Audit statistics and trends
GET
/api/v1/skills
—
List all audited skills
POST
/api/v1/feedback
—
Submit TP/FP feedback
POST
/api/v1/intel/submit
—
Submit threat intelligence
GET
DOCS
/api/v1/attack/technique/{id}
—
MITRE ATT&CK technique lookup
/api/v1/docs
—
Interactive Swagger docs →
Authentication
All audit requests require an X-API-Key header. Get a key via:
- Browser: Click Login (GitHub OAuth) → key auto-saved
- Agent: POST
/api/v1/api-keys/register→ GitHub Device Flow → poll/verify
Rate limit: 100 requests/week per key.
Contact
Get in Touch
Questions, feedback, or partnership inquiries about ClawSec skill security audit service.
huacheng99@gmail.com