Security audit

Agent群组工具免费版

Security checks for vulnerabilities and agentic risk

Overview

This appears to be a local group-agent coordination skill, with some overbroad boilerplate that users should keep scoped to the intended CLI workflow.

Install only if you intend to use a local group-agent CLI for agent group coordination. Keep use limited to the documented group-agent commands, review where SQLite data and logs are stored, and do not allow the broad file/API/command wording to justify unrelated shell automation without explicit user approval.

SkillSpector

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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

Findings (5)

Intent-Code Divergence

Medium

Confidence: 90% confidence
Finding: The skill repeatedly presents itself as a local-only, single-instance free edition with no webhook or cross-instance integrations, but later introduces network-oriented guidance such as HTTPS usage, ping-based connectivity checks, and generic external API framing. This inconsistency can cause an agent or operator to over-trust the skill's boundaries and enable or troubleshoot network behavior that was supposedly out of scope, weakening safe-use assumptions.

Description-Behavior Mismatch

Medium

Confidence: 94% confidence
Finding: The file's later sections broaden the skill from a narrowly scoped group-collaboration tool into a generic automation capability covering file handling, API integration, and command execution. Scope inflation is dangerous because orchestrators may invoke the skill for unrelated tasks and grant it read/exec authority under a misleading description, increasing the chance of unsafe actions outside user intent.

Vague Triggers

Medium

Confidence: 87% confidence
Finding: The trigger condition says to use the skill whenever project management, planning, progress tracking, or collaboration is needed, which is broad enough to match many ordinary requests. Over-broad routing criteria can cause this read/exec-capable skill to be auto-selected in situations where simple discussion or planning was intended, unnecessarily expanding execution privileges.

Vague Triggers

Medium

Confidence: 86% confidence
Finding: The manifest description advertises broad applicability across project management, task planning, progress tracking, team collaboration, independent developers, enterprise teams, and automation workflows without strong invocation limits. In a skill ecosystem, this kind of expansive wording can attract the skill into many contexts, increasing the likelihood that read/exec tools are used when a narrower non-executing skill would suffice.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The skill claims file processing, command execution, and automated workflows but does not give a prominent user-facing warning that these actions may modify local files or system state. When paired with allowed read/exec tools, this omission increases the risk of unintended side effects because users and orchestrators may not realize the skill can perform impactful operations beyond passive coordination.

Static analysis

No suspicious patterns detected.