ClawHub
Back to skill

Security audit

Free Resource Finder Free

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed helper for finding and switching free AI model providers, with some privacy and persistence caveats users should understand.

Install only if you intend to let the agent use an external CLI to change model configuration and contact third-party model providers. Verify the free-finder package source, use a scoped API key, avoid sending secrets or regulated data to free model providers, and back up existing model configuration before running switch or reset commands.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Vague Triggers

Medium
Confidence
86% confidence
Finding
The skill declares an extremely broad keyword trigger scope with many generic phrases and no meaningful activation constraints. In an agent environment with read/write/exec permissions, overbroad triggering can cause the skill to activate unexpectedly and steer workflows into configuration changes or external API usage without clear user intent.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill describes model-switching commands that modify active configuration, but the warning about local config updates appears only later in an example rather than as a clear upfront safety notice. In a tool with write and exec capabilities, hidden or underemphasized state changes can lead users or agents to unintentionally alter persistent settings and disrupt other workflows.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The API key setup and usage instructions do not clearly warn that prompts, metadata, and possibly sensitive project content will be sent to a third-party API provider. Because this skill is specifically designed to reroute model usage through external services, missing privacy disclosure increases the risk of accidental data exfiltration from users who assume processing is local or first-party.

Static analysis

No suspicious patterns detected.