briefing-visualizer简报转手机长图

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:45 PM UTC
• Version:
  1.0.0

Overview

This skill appears to do what it says: create a mobile-friendly briefing image locally, with normal caution needed because it runs a Python script and headless Chrome.

Install only if you are comfortable running a local Python script that launches Chrome. Use trusted briefing content and banner images, avoid confidential or unauthorized scraped material, and review generated HTML before rendering when it includes external links, scripts, or third-party assets.

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 Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands

Findings (4)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 88% confidence
• Finding: The skill instructs the agent to execute a local shell command (`python3 process_briefing.py ...`) and to use Chrome headless for rendering, but the skill declares no permissions for shell/code execution. Even though this appears intended for legitimate image generation, undeclared execution capability is dangerous because user-controlled file paths and content could lead to unsafe command invocation, unexpected local file access, or execution in environments that rely on permission declarations for sandboxing and review.

Description-Behavior Mismatch

Medium

• Confidence: 97% confidence
• Finding: The skill advertises transformation of structured briefing content, but the implementation loads arbitrary local HTML into a full Chrome engine. Rendering untrusted HTML can execute active content, fetch remote resources, and abuse browser file/origin behavior, making the capability substantially broader and riskier than the stated purpose.

Context-Inappropriate Capability

Medium

• Confidence: 93% confidence
• Finding: Launching an external browser subprocess to process user-controlled content expands the attack surface beyond simple image generation. Even without shell injection, a real browser may perform network access, load external assets, and expose the host to browser-engine risks or unintended local file rendering behaviors.

Missing User Warnings

Medium

• Confidence: 84% confidence
• Finding: The publish guide advertises 'AI-assisted content fetching' as a core capability but does not describe what sources may be accessed, what user data may be transmitted, or what consent and validation controls exist. In a skill that processes user-provided briefing content, this omission can lead to unexpected network access, leakage of sensitive business information, or fetching untrusted external content into the workflow.

VirusTotal

66/66 vendors flagged this skill as clean.