ClawHub
Back to skill

Security audit

meta-toolchain-orchestrator

Security checks for vulnerabilities and agentic risk

Overview

This skill is not clearly malicious, but it asks users to trust broad orchestration, self-verification, memory, and self-evolution behavior that the artifact does not adequately implement or scope.

Review this skill before installing. Treat it as an experimental generated meta-skill, not a verified orchestrator. Do not allow it to run or repair scripts, persist task notes, or participate in long-running agent loops unless you first define explicit command, path, logging, and memory limits.

SkillSpector

Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
73% confidence
Finding
The skill advertises operational capabilities such as logging, learner integration, and persistence-related behavior, but declares no explicit permissions. In an agent ecosystem, undeclared file-write capability weakens trust boundaries and can enable unexpected local state mutation, artifact tampering, or persistence of sensitive task data without informed approval.

Tp4

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding
This skill makes strong claims about self-verification, self-reflection, orchestration, and continuous self-evolution, but the described implementation does not substantiate those controls and instead includes persistent local storage. Security-relevant overclaiming is dangerous because operators may rely on nonexistent safeguards, causing unsafe automation decisions, reduced oversight, and unjustified trust in outputs or stored state.

Static analysis

No suspicious patterns detected.