ClawHub
Back to skill

Security audit

api-gateway

Security checks for vulnerabilities and agentic risk

Overview

The skill is a disclosed API gateway, but its sensitive local caching behavior is broader and less consistently described than users are told.

Review the local data handling before installing. Use environment variables where possible, avoid sending sensitive or regulated data through it, avoid --cache-full for sensitive providers, and clear cache/log files after use. Treat memory/api-gateway/cache.json and keys.json as sensitive local files.

SkillSpector

Vulnerability Patterns
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Intent-Code Divergence

Medium
Confidence
90% confidence
Finding
The document gives conflicting statements about what is persisted: earlier sections say request logs omit endpoints and bodies and cache is metadata-only by default, while the later security notes say logs/cache may persist request bodies, endpoints, and response headers. This inconsistency can cause operators to make unsafe assumptions about local data exposure and may lead them to send sensitive data believing it will not be stored.

Missing User Warnings

Medium
Confidence
84% confidence
Finding
The README explicitly advertises request logging and response caching for API calls, but it does not disclose that these features may store sensitive request bodies, headers, tokens, or API responses. In an API gateway context, that omission can lead operators to enable or use the skill without realizing secrets or regulated data may be retained in logs or cache, increasing the risk of unintended exposure.

Static analysis

No suspicious patterns detected.