ClawHub
Back to skill

Security audit

个人效率顾问基础版

Security checks across malware telemetry and agentic risk

Overview

This productivity skill is mostly a Markdown guide, but it asks for broad command/API capability and gives inconsistent privacy and credential-handling instructions.

Review this before installing. It may be acceptable if you only use it as advisory Markdown, but do not let it run shell commands or inspect environment variables unless you have checked the exact command. Treat its local-only privacy claim as unreliable until the publisher clearly explains what data can be sent to external APIs and when user confirmation is required.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Description-Behavior Mismatch

High
Confidence
97% confidence
Finding
The skill states that FREE-version data is stored locally and not uploaded to the cloud, but elsewhere it explicitly allows network connectivity and external API usage. This creates a misleading trust boundary: users may disclose sensitive productivity notes, schedules, or personal data under the assumption that nothing leaves the device when the agent may in fact transmit data externally.

Context-Inappropriate Capability

Medium
Confidence
95% confidence
Finding
The skill instructs validating configuration by enumerating environment variables matching API, KEY, TOKEN, or SECRET, which exposes the presence and naming of sensitive credentials unrelated to a personal productivity assistant. Even with value masking, this capability encourages credential discovery and expands the agent's access to secrets beyond what is necessary for the stated function.

Intent-Code Divergence

High
Confidence
98% confidence
Finding
The documentation claims all FREE-version data remains local and is never uploaded, while nearby sections describe dependence on external APIs and network access. This contradiction can cause users to provide sensitive personal information based on false privacy assurances, making accidental exfiltration or policy noncompliance more likely.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill says the agent will automatically call tools or APIs and execute operations based on natural-language requests, but it does not provide clear warnings, consent boundaries, or approval steps for command execution and external calls. In an agentic environment with exec enabled, this increases the risk of unreviewed shell commands, unintended network requests, or unsafe actions triggered from ordinary user prompts.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.