Security audit

Tmp.Fm4eKnkduc

Security checks across malware telemetry and agentic risk

Overview

This skill is a legitimate skill-management helper, but it can broadly scan and modify local skills and may publish externally, so users should review its scope before installing.

Install only if you want an agent to manage your local skill library. Before using it, confirm each scan, local SKILL.md edit, and publication target explicitly, especially for GitHub, ClawHub, or Xiaping publishing.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
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

Findings (5)

Context-Inappropriate Capability

Medium

Confidence: 90% confidence
Finding: The instructions require enumerating the entire local skills directory for every new skill request, even when unnecessary. That broad discovery behavior increases exposure of local metadata and repository structure and goes beyond what is minimally needed for many create/publish tasks.

Description-Behavior Mismatch

Medium

Confidence: 76% confidence
Finding: The manifest says the skill triggers for install and publish requests, but the workflow always performs overlap assessment and file creation/upgrade first. This can cause unintended local modifications during requests that users may reasonably expect to be limited to installation or publishing, increasing the risk of accidental changes.

Vague Triggers

High

Confidence: 94% confidence
Finding: The trigger phrases are extremely broad and likely to match ordinary requests about making or evaluating a skill. In context, this skill can read local directories, write SKILL.md files, and publish externally, so overbroad invocation materially raises the chance of unintended activation of sensitive capabilities.

Vague Triggers

High

Confidence: 93% confidence
Finding: The description states the skill applies to creation, installation, publishing, or updating of any skill, which is too ambiguous for a capability that can modify the local skill store and perform network publication. Such broad scope makes accidental or unjustified invocation more likely and weakens the principle of least astonishment.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The description does not clearly warn users that the skill may modify local skill files and publish content to external platforms. For a skill with file-write, shell, and network behaviors, missing this warning undermines informed consent and increases the chance of unintended data exposure or unauthorized publication.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.