Back to skill

Security audit

数字人口播视频工作室|AI-HIVE

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly aligned with AI-HIVE video generation, but its script can expose the API key to an arbitrary configured server and can write downloaded results outside the chosen folder if the API response is malicious.

Install only if you are comfortable reviewing commands before execution. Use the default AI-HIVE endpoint, avoid setting --base-url or AI_HIVE_BASE_URL unless you fully trust it, consider a limited or throwaway API key, and run downloads in a controlled directory.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (2)

T09 · Insecure Skill Coding Practices

Error
Location
scripts/videogen.py:91
Finding

API Credential Disclosure Through an Unrestricted Base URL

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
scripts/videogen.py:397
Finding

Arbitrary File Write Through Unsanitized Download Paths

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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
Findings (8)

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

This second mismatch finding indicates the skill exposes broader capabilities such as generic chat/image generation, account or wallet queries, model enumeration, and interactive API-key setup, while omitting key declared controls like script generation structure, subtitle extraction, fact-checking, manual review, and authorization handling. In a content-production skill that can touch external APIs and potentially billable resources, undeclared capability expansion materially raises the risk of secret misuse, surprise charges, privacy issues, and bypass of expected review gates.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

This second mismatch finding indicates the skill exposes broader capabilities such as generic chat/image generation, account or wallet queries, model enumeration, and interactive API-key setup, while omitting key declared controls like script generation structure, subtitle extraction, fact-checking, manual review, and authorization handling. In a content-production skill that can touch external APIs and potentially billable resources, undeclared capability expansion materially raises the risk of secret misuse, surprise charges, privacy issues, and bypass of expected review gates.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
88% confidence
Finding

The skill advertises runnable commands and operational behaviors that imply access to environment variables, local files, networking, and shell execution, but it declares no explicit tool scope or permissions boundary. This creates a capability-transparency problem: a host agent may execute the skill with broader privileges than users or reviewers expect, increasing the chance of secret exposure, unintended file access, or networked actions such as billable API calls.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The natural-language prompt explicitly instructs the skill to produce a "production-ready Chinese workflow," which imposes a language choice by default. This is a language/locale policy concern because the file does not indicate that users can opt in to Chinese or choose another language.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The file materially exceeds the declared skill scope. Although marketed as a digital-human talking-video studio, it also exposes generic chat, image generation, model enumeration, account inspection, and upload primitives, which broadens the accessible attack surface and enables off-purpose use beyond the user’s expected authorization boundary. In agent environments, this kind of capability drift is dangerous because downstream orchestration may grant or invoke the skill under narrower assumptions than the code actually implements.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

This code file contains natural-language descriptions and user-facing messages exclusively in Chinese, including the module docstring and later CLI help/output strings. Under the stated policy, forcing a specific language without user opt-in is a natural-language policy violation unless a locale restriction is clearly documented and justified, which is not present here.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
80% confidence
Finding

This code embeds the skill name and command description entirely in Chinese, which can impose a specific language experience on users without any opt-in or alternative locale. The policy explicitly calls for flagging language or locale constraints when the skill does not offer a user choice or clearly justify the restriction.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

The exposed user-info and models commands provide account and platform-inspection functionality not necessary for generating digital-human videos. While not directly code-execution dangerous, they reveal extra operational data and can facilitate reconnaissance about account state, available models, routing modes, and service configuration that an agent or user did not need to accomplish the stated task.

Content

No source excerpt is available for this finding.