Security audit

效率倍增器

Security checks across malware telemetry and agentic risk

Overview

The skill is a coherent local productivity workflow, but it asks for shell execution while telling users it does not need exec, so it should be reviewed before installation.

Install only if you are comfortable with a local productivity skill that may create and reorganize files under ~/productivity/. Review the exec permission first; for this Markdown-only workflow it should be removed or explicitly justified with narrow command rules before routine use.

SkillSpector

By NVIDIA

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 (6)

Intent-Code Divergence

Medium

Confidence: 97% confidence
Finding: The skill claims it is 'pure Markdown' and '无需 exec', yet the manifest exposes the exec tool. This mismatch can cause reviewers and users to underestimate the skill's capabilities, and if the agent or surrounding platform permits tool use based on the manifest, the skill may gain shell execution far beyond what the documentation suggests.

Description-Behavior Mismatch

Medium

Confidence: 96% confidence
Finding: The skill is presented as a local file-based productivity workflow, but it declares shell execution capability that is not necessary for the described behavior. This expands the attack surface from simple file manipulation to arbitrary command execution, which could enable destructive local actions if the skill is invoked in unexpected contexts or later modified.

Context-Inappropriate Capability

Medium

Confidence: 95% confidence
Finding: Shell execution is unjustified for productivity planning, note-taking, and local Markdown organization as described in this skill. Unnecessary privileged capability violates least privilege and increases the chance that prompt injection, misrouting, or future instruction changes could turn a benign planning skill into a vehicle for arbitrary local command execution.

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The trigger keywords are very broad, including common terms like '效率', '生产力', '计划', 'focus', and 'productivity'. This makes accidental invocation more likely, which is more concerning here because the skill can read and potentially modify local productivity files and currently also declares exec.

Vague Triggers

Medium

Confidence: 86% confidence
Finding: The activation guidance is framed as broad situations rather than clear boundaries, so the skill may be selected in loosely related conversations about work stress, planning, or focus. Ambiguous routing increases the chance of unintentional file operations or exposure of local planning data when the user did not specifically ask for this system.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The skill includes instructions to create, rewrite, and reorganize local task files, but the main usage guidance does not prominently warn that existing user data may be modified. Users may assume they are receiving planning advice only, while the agent could overwrite or restructure files, causing data loss, confusion, or unwanted persistence.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.