ClawHub
Back to skill

Security audit

Alpine Linux

Security checks for vulnerabilities and agentic risk

Overview

This skill only fetches and summarizes public Alpine Linux release information, with no evidence of hidden data access, persistence, or destructive behavior.

Install only if you are comfortable with the skill running local shell scripts that contact Alpine Linux release servers and parse their release metadata. Review or constrain those scripts in locked-down environments, and note that the metadata should more clearly advertise shell and network use.

SkillSpector

Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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 (1)

Lp3

Medium
Category
MCP Least Privilege
Confidence
90% confidence
Finding
The skill invokes shell scripts (`bash scripts/alpine_fetch_latest_releases.sh` and `bash scripts/alpine_aggregate_latest_version_archs.sh`) but does not declare corresponding permissions. This creates a capability/permission mismatch that can bypass operator expectations and weakens security review because the skill can execute code despite appearing less privileged.

Static analysis

No suspicious patterns detected.