Back to skill

Security audit

第一视角POV商品广告|AI-HIVE

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly a disclosed AI-HIVE product-ad video workflow, but it has a real credential-safety concern because its API key can be sent to an arbitrary configured API base URL.

Install only if you trust the AI-HIVE workflow and can control how it is invoked. Do not set AI_HIVE_BASE_URL or config-file base_url to anything except the official AI-HIVE endpoint, avoid passing API keys on shared command lines, and use only media files you trust for concat/editing. Review parameters before any generation call because API use may be billable.

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:90
Finding

AI-HIVE API Key Disclosure Through Unrestricted Base URL Override

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/edit_video.py:78
Finding

FFmpeg Concat Manifest Injection Through Crafted Input Filenames

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • 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
Findings (12)

Tp4

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

This second mismatch indicates the skill is presented as a creative-generation and AI-HIVE workflow tool, but in practice it performs undeclared local media-processing operations via ffmpeg/ffprobe while lacking the promised ad-planning and API behaviors. Such hidden or shifted functionality increases risk because users may authorize the skill for one purpose while it exercises different local-processing capabilities, weakening transparency and informed consent in an automated environment.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

This second mismatch indicates the skill is presented as a creative-generation and AI-HIVE workflow tool, but in practice it performs undeclared local media-processing operations via ffmpeg/ffprobe while lacking the promised ad-planning and API behaviors. Such hidden or shifted functionality increases risk because users may authorize the skill for one purpose while it exercises different local-processing capabilities, weakening transparency and informed consent in an automated environment.

Content

No source excerpt is available for this finding.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 81)May include surrounding context.

md
python3 scripts/videogen.py init --skill-name pov-product-ad-ai-hive

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 83)May include surrounding context.

md
python3 scripts/videogen.py init --skill-name pov-product-ad-ai-hive

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
93% confidence
Finding

The skill advertises and demonstrates code paths that use environment variables, local file access, shell commands, and network access, but it does not declare any tool scope or permission boundaries. In an agent setting, this can lead to overbroad execution privileges, making unintended filesystem access, command execution, or outbound requests easier if the skill is invoked automatically or reused in a broader workflow.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The default prompt hard-codes Chinese output ('production-ready Chinese workflow') regardless of the user's stated preference. This can override user intent, reduce transparency, and create prompt-steering behavior that is especially problematic in agent chains where downstream tools or reviewers assume the model will preserve user language and instructions.

Content

No source excerpt is available for this finding.

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/edit_video.py (reported line 20)May include surrounding context.

python
def run(args: list[str]) -> None:
    print(" ".join(args))
    subprocess.run(args, check=True)


def main() -> None:

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The module docstring and the code expose broad chat, image generation, model listing, user-info, generic upload, and arbitrary video-generation capabilities, effectively acting as a reusable AI Hive CLI rather than a narrowly scoped POV商品广告 skill. The manifest describes a specialized skill for first-person product-ad workflows, while this file provides a much wider set of AI operations unrelated to that constrained purpose.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The top-level docstring says this is a '通用 AI 能力调用工具' wrapping text chat, image generation, video generation, model query, media upload, and polling. However, the embedded skill configuration at L775-L776 identifies a specific POV商品广告 skill, so the documentation actively frames the code as a generic tool rather than the specialized skill it claims to be.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

This code file contains natural-language descriptions, help text, and user-facing messages predominantly in Chinese, starting from the module docstring. The policy allows locale constraints only when users are given a choice or when the constraint is clearly documented and justified; neither is present here.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The docstring and argparse description are entirely in Chinese, and the script hardcodes Chinese skill metadata values. This creates a language-specific user experience without offering any language or locale choice, which matches the stated policy concern for forced language use.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

The get_user_info capability and corresponding user-info command retrieve user information and wallet balance, which is not needed to generate first-person product-ad scripts or videos. This is an extra operational/account-management capability not justified by the manifest’s stated creative-production purpose.

Content

No source excerpt is available for this finding.