ClawHub
Back to skill

Security audit

Fly Install

Security checks for vulnerabilities and agentic risk

Overview

This skill is an advertised fallback installer, but it can automatically install unverified third-party skills into the agent's skills directory.

Use this only when you already trust the exact skill source. Before running it, verify the GitHub owner/repo or ClawHub download page yourself, avoid batch installs, avoid path-like skill names or broad custom directories, and review the installed SKILL.md before letting an agent use it.

SkillSpector

Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The script fetches code and content from GitHub or a download API and installs it directly into the user's skills directory with only a minimal existence check for SKILL.md. That creates a supply-chain trust problem: a malicious or spoofed repository/zip can be imported and later executed or consumed by the agent environment without any provenance check, signature verification, publisher allowlist, or explicit user warning about untrusted content.

Static analysis

No suspicious patterns detected.