ClawHub
Back to skill

Security audit

English-Learning

Security checks for vulnerabilities and agentic risk

Overview

This skill is a simple English listening-practice helper that fetches lesson content from Supabase, with a credential-hygiene caution but no evidence of hidden or harmful behavior.

Before installing, replace the example Supabase values with your own environment variables, rotate the shown anon key if it was ever real, and confirm the Supabase anon role can only read intended public listening-material tables and cannot write or access private data.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Context-Inappropriate Capability

Medium
Confidence
96% confidence
Finding
The skill documentation includes a concrete Supabase anon key example directly in the file, which normalizes embedding live-looking credentials in distributable documentation. Even if labeled as an example, such tokens are often real, reusable, or copied into deployments, and they encourage unsafe secret-handling practices unrelated to the educational purpose of the skill.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The skill instructs users to send a Supabase anon key in API requests but provides no warning about data exposure, tenant scoping, or the fact that all fetched listening content and metadata are transmitted to an external backend. This is risky because operators may assume the key is harmless and deploy without verifying Row Level Security, least-privilege access, or privacy implications of the returned data.

Static analysis

No suspicious patterns detected.