UAPI 生成摸摸头GIF (QQ号) 接口

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:17 PM UTC
• Version:
  1.0.0

Overview

This skill is a documentation-only wrapper for one public image API, but its keywords are broader than the actual feature it provides.

Install this only for generating a motou GIF from a QQ number. Treat the generic image-processing keywords as inaccurate, and avoid using it for compression, conversion, SVG rendering, or NSFW detection. Only provide QQ numbers or a UAPI key if you are comfortable sending them to the UAPI service.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (5)

Description-Behavior Mismatch

Medium

• Confidence: 94% confidence
• Finding: The skill claims to wrap only GET /image/motou, but its manifest also advertises unrelated capabilities like image compression, image/base64 conversion, and other image tasks. This can cause the orchestrator to route requests for broader image operations into a skill that is not scoped for them, increasing the chance of unintended API calls, misuse, or user-task confusion.

Intent-Code Divergence

Medium

• Confidence: 95% confidence
• Finding: The keyword section contradicts the documented single-endpoint purpose by listing unrelated operations such as image compression, base64 conversion, SVG-to-image, and NSFW image detection. In a skill-selection system, these misleading keywords can cause invocation for tasks the skill cannot safely or correctly fulfill, undermining routing integrity.

Description-Behavior Mismatch

Medium

• Confidence: 95% confidence
• Finding: The skill claims to wrap only GET /image/motou, but its keyword index advertises unrelated capabilities such as image compression, image/base64 conversion, SVG to image, and NSFW detection. This can cause incorrect tool selection by an agent, leading to unintended API calls, confused-deputy behavior, or mishandling of user requests outside the skill's actual scope.

Vague Triggers

Medium

• Confidence: 92% confidence
• Finding: The English trigger phrases in the manifest are overly broad and include generic image-processing terms unrelated to the endpoint. This broad matching can unintentionally activate the skill for inappropriate requests, leading to incorrect downstream behavior and possible policy bypass in systems that rely on textual routing.

Vague Triggers

Medium

• Confidence: 96% confidence
• Finding: The keyword list includes ambiguous and unrelated triggers without any scope guardrails, including terms tied to unrelated image functions. Because skill dispatch often depends on keyword overlap, this creates a practical risk of accidental or overbroad invocation for user intents outside the skill's true functionality.

VirusTotal

64/64 vendors flagged this skill as clean.