MiniMax Search & Vision

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:26 PM UTC
• Version:
  1.0.0

Overview

The skill coherently provides MiniMax-backed web search and image analysis, with normal external-service privacy cautions but no evidence of hidden or destructive behavior.

Install only if you trust MiniMax, mcporter, and the MCP package it runs. Use a scoped MiniMax API key, keep the credential file private, and avoid sending confidential searches, internal URLs, screenshots, or local images unless external MiniMax processing is acceptable.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (6)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 87% confidence
• Finding: The skill describes capabilities that involve reading environment variables, accessing credential files, and invoking shell commands via mcporter, but it does not declare permissions or boundaries for those actions. This can lead to overbroad execution expectations and makes it easier for an agent framework to invoke sensitive capabilities without clear user awareness or policy gating.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The README advertises web search and image understanding features but does not clearly disclose that user prompts, image URLs, and potentially local image contents will be sent to an external MiniMax service. This can lead users to submit sensitive data under the mistaken assumption processing is local, creating a privacy and data-handling risk in a tool explicitly designed to transmit content off-host.

Vague Triggers

Medium

• Confidence: 84% confidence
• Finding: The trigger phrases are broad and overlap with common user requests such as general searching or image explanation, which increases the chance the skill activates unintentionally. Because the skill can cause external network access and image transmission, accidental invocation may send user data to a third-party service without sufficiently explicit intent.

Missing User Warnings

Medium

• Confidence: 93% confidence
• Finding: The skill enables web search and remote image analysis but does not clearly warn that user queries, image URLs, or local file paths may be transmitted to an external service. This creates a privacy and data handling risk, especially if users provide sensitive images or local paths without understanding the data leaves the local environment.

Missing User Warnings

Medium

• Confidence: 93% confidence
• Finding: The examples encourage users to submit both remote image URLs and local file paths to an image-understanding tool without any disclosure that the image content, file path, and possibly related metadata may be transmitted to an external MiniMax service. In a tool specifically designed for web search and image analysis, this omission increases the chance of unintentional disclosure of sensitive local screenshots, internal URLs, or confidential images by users who may assume processing is local.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The documentation explicitly allows `image_url` to be either a remote URL or a local file path, but it does not warn users that the referenced content will be transmitted to an external MiniMax service for analysis. In a skill designed for web search and image understanding, this omission increases the chance that users or downstream agents may unintentionally upload sensitive local files, internal screenshots, or intranet-accessible resources to a third party, creating a real privacy and data-exfiltration risk.

VirusTotal

66/66 vendors flagged this skill as clean.