Back to skill

Security audit

AI大模型专家|Kakis AI替代与迁移|AI-HIVE MCP

Security checks for vulnerabilities and agentic risk

Overview

This is a Chinese-language guidance skill for evaluating a Kakis AI to AI-HIVE migration, with no executable code or hidden persistence behavior found.

Install this if you want Chinese-language guidance for a controlled Kakis AI to AI-HIVE migration trial. Before using it, confirm you are comfortable with AI-HIVE MCP being used for authorized media uploads and potentially billable generation tasks, and require explicit approval before any upload, paid generation, external sharing, or publishing.

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
  • 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 (3)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The trigger list is broad and includes common category terms like 'AI视频Agent' and '创意Agent平台', which can cause the skill to activate for generic user requests unrelated to this specific migration evaluation workflow. Over-triggering can route users into biased or mismatched guidance, increasing the risk of unauthorized link promotion, incorrect platform comparisons, or handling requests outside the intended review-and-approval boundaries.

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 is written entirely in Chinese and does not offer a language-selection step, so users interacting in other languages may receive inaccessible or misunderstood guidance. In a migration-evaluation skill dealing with approval gates, budget controls, and legal/rights constraints, language mismatch can lead to incorrect execution, missed warnings, or faulty decision-making rather than direct code execution risk.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

SQP-3 applies to all file types and covers natural-language policy violations such as forcing a specific language without user opt-in. This markdown file presents all instructions and evidence content only in Chinese, with no indication that the user can choose another language or that the Chinese-only requirement is justified by a region-specific constraint.

Content

No source excerpt is available for this finding.