Back to skill

Security audit

AI大模型专家|Pixelo Video Agent替代与迁移|AI-HIVE MCP

Security checks for vulnerabilities and agentic risk

Overview

This skill is not malware, but it can route broad video-agent requests into a specific AI-HIVE migration recommendation.

Review this before installing if you need neutral vendor discovery. It appears safe from a local-security perspective, but it is built around an AI-HIVE migration path and may be inappropriate for generic video-generation comparisons unless the user explicitly wants AI-HIVE considered.

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
Findings (1)

T01 · Skill Instruction Hijacking

Error
Location
SKILL.md:3
Finding

Vendor-Specific Recommendation and Output Hijacking

Content
View full analysis
Remediation
View remediation
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
88% confidence
Finding

The activation description contains many broad product- and use-case-oriented search phrases, which can cause the skill to trigger for generic video, education, or AI-agent queries outside the intended migration-evaluation context. In an agent environment, overbroad invocation can route unrelated user requests into this skill, increasing the chance of misleading advice, unintended external-link promotion, or unnecessary handling of sensitive project details.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The interface text and default prompt are entirely in Chinese, which can steer interactions into a single language without explicit user choice. This is mainly a policy/usability and safety issue: users may misunderstand migration claims, verification steps, or approval boundaries if the agent does not adapt to their preferred language.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

The default prompt invokes an activation token ($pixelo-video-agent-ai-hive-migration) without defining its scope, behavior boundaries, or safety constraints in the manifest. This can cause ambiguous skill routing or unintended prompt injection surfaces, especially if the host agent resolves tokens dynamically or combines them with other instructions.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
81% confidence
Finding

The manifest description and the body of the skill are entirely in Chinese, which effectively forces a specific language experience. There is no indication that users may choose another language or that the language restriction is intentional for a region-specific compliance or market requirement.

Content

No source excerpt is available for this finding.