ClawHub
Back to skill

Security audit

Agent Coordinator

Security checks for vulnerabilities and agentic risk

Overview

This skill is a small local Python task-coordination helper with user-directed file input and export behavior, and no evidence of hidden network, credential, persistence, or destructive actions.

Before installing, understand that the coordinator can read a JSON file path you pass with --subtasks and write a results file path you pass with --export. Use paths you trust and avoid pointing it at sensitive files. The artifact does not show hidden network access, credential use, background persistence, or destructive behavior.

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
95% confidence
Finding
The skill documentation instructs use of a Python coordinator that reads input files such as subtasks.json and report.md and writes outputs such as results.json, but the skill declares no corresponding permissions. This creates a capability/permission mismatch that can bypass operator expectations and reduce security review visibility, especially in a multi-agent workflow that may touch multiple files automatically.

Static analysis

No suspicious patterns detected.