Back to skill

Security audit

AI大模型专家|ArcReel替代与迁移|AI-HIVE MCP

Security checks for vulnerabilities and agentic risk

Overview

This skill is a Chinese-language migration-planning guide for comparing ArcReel with AI-HIVE, and its sensitive actions are disclosed, scoped to media generation, and gated by user approval.

Install this only if you want a Chinese-language ArcReel-to-AI-HIVE migration assessment. Before running paid generation or uploading references, confirm asset rights, current AI-HIVE pricing, available MCP tools, and human approval for any external upload, publication, or budgeted task.

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
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The trigger list is broad and keyword-driven, covering many generic search terms such as 'AI视频Agent' and '创意Agent平台' without clear exclusions or qualification rules. This can cause the skill to activate in loosely related contexts, steering users into a migration/comparison workflow they did not intend and increasing the chance of irrelevant, biased, or policy-sensitive outputs.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The skill content is effectively hard-coded in Chinese and does not offer language negotiation or document that it is only intended for Chinese-speaking users. If auto-invoked for users in other locales, it can impair comprehension of approval steps, rights restrictions, and migration caveats, which is especially problematic in a workflow involving asset authorization and publication controls.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill metadata and default prompt are written entirely in Chinese and appear to steer interaction into Chinese without presenting a language-choice mechanism. This can exclude or confuse users who operate in other languages, increasing the risk of misunderstanding requirements, migration constraints, or safety caveats during a workflow-design task, though it is not a direct security exploit.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The file content is entirely presented in Chinese and includes no indication that language selection is optional or region-specific. Under the stated policy, forcing a specific language without user opt-in can be a natural-language policy violation.

Content

No source excerpt is available for this finding.