JavaScript开发工具

• Outcome: Review
• Latest audit: Aug 10, 2026 · 11:35 AM UTC
• Version: 1.0.1

Overview

This JavaScript helper is not clearly malicious, but it asks for command and file-writing power while documenting broad, inconsistently scoped automation features.

Review before installing. Use this only if you are comfortable granting a JavaScript assistant read, write, and command execution authority, and keep commands and file changes explicitly user-directed. The artifact does not show exfiltration or hidden persistence, but its permissions and automation claims are broader than its JavaScript-focused description.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (5)

Description-Behavior Mismatch

Medium

• Confidence: 93% confidence
• Finding: The skill is presented as a JavaScript coding/review assistant, but the documentation expands its scope to file processing, API integration, search, and command execution. This capability creep weakens least-privilege expectations and can mislead users or agents into granting broader actions than are necessary for the stated purpose.

Context-Inappropriate Capability

Medium

• Confidence: 95% confidence
• Finding: Advertising system command execution for a JavaScript writing/review skill is dangerous because exec enables arbitrary OS-level actions unrelated to code review. If an agent follows the skill literally, this can lead to unauthorized command execution, filesystem changes, or chaining into broader compromise.

Context-Inappropriate Capability

Low

• Confidence: 82% confidence
• Finding: The skill claims API integration and information retrieval capabilities that are not clearly necessary for its JavaScript-focused purpose. While less severe than exec, unnecessary network and retrieval capabilities expand the attack surface and can enable unintended data access or exfiltration in agent workflows.

Intent-Code Divergence

Medium

• Confidence: 90% confidence
• Finding: The document inconsistently describes the skill as both pure Markdown/natural-language driven and execute-capable. This discrepancy can cause users or orchestrators to underestimate the skill's privilege level, increasing the likelihood that dangerous execution features are enabled without appropriate scrutiny.

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The skill advertises file writing, external API calls, and command execution without strong user-facing warnings about side effects or security impact. In an agent setting, missing warnings reduce informed consent and make it easier for high-impact actions to be triggered under the guise of routine JavaScript help.

VirusTotal

VirusTotal findings are pending for this skill version.

Static analysis

No suspicious patterns detected.