Security audit

aicreat

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward NocoBase administration guide for creating and managing AI employees, with disclosed configuration-changing abilities but no hidden code or deceptive behavior.

Install this only if you intend to let the agent help administer NocoBase AI employees and page integrations. Review proposed create, update, delete, shortcut, and button changes before applying them, and keep AI employee data scopes and tool bindings limited to the tables and workflows they actually need.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

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
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (5)

Tp2

High

Category: MCP Tool Poisoning
Confidence: 85% confidence
Finding: Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.
Content: No source excerpt is available for this finding.

Tp2

High

Category: MCP Tool Poisoning
Confidence: 85% confidence
Finding: Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.
Content: No source excerpt is available for this finding.

Tp2

High

Category: MCP Tool Poisoning
Confidence: 85% confidence
Finding: Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The trigger list includes broad terms such as "chatbot", "头像", and "浮动按钮", which can match many unrelated user requests and cause this skill to activate outside its intended administrative context. Because the skill can guide or invoke creation and modification of AI employees and page integrations, unintended activation increases the chance of accidental configuration changes or exposing management workflows when the user only wanted general help.
Content: No source excerpt is available for this finding.

Overly Broad Trigger

Low

Category: Trigger Abuse
Confidence: 70% confidence
Finding: Overly Broad Trigger: '头像' is too short and may match unintended inputs
Content: No source excerpt is available for this finding.