Back to skill

Security audit

美妆步骤教程短视频|AI-HIVE

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly matches its AI-HIVE beauty-video purpose, but it handles a billable API key with insufficient endpoint scoping and ships broader client code than the public workflow needs.

Review before installing. Use this only with a dedicated AI-HIVE key that has limited quota, keep the default AI-HIVE endpoint unless you intentionally trust another endpoint, avoid setting AI_HIVE_BASE_URL from untrusted environments, and do not upload media unless you have the right to use it. Prefer installing dependencies in a virtual environment with pinned versions.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
Findings (2)

T08 · Insecure Dependencies

Warning
Location
scripts/videogen.py:14
Finding

Unpinned Runtime Dependency Installation

Content
View full analysis
Remediation
View remediation
`. 2. Generate and enforce cryptographic hashes using a locked requirements file and `pip install --require-hashes -r requirements.txt`. 3. Document a trusted HTTPS package index rather than relying silently on the user's active pip configuration. 4. Use a dependency-locking workflow such as `pip-tools`, Poetry, or an equivalent reproducible build mechanism. 5. Add automated dependency vulnerability and provenance scanning to the release process. 6. Recommend installation inside a dedicated virtual environment with no unnecessary credentials or filesystem privileges. 7. Replace the generic installation instruction with the exact, integrity-checked installation command supported by the project. ]]>

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/videogen.py:92
Finding

AI-HIVE Bearer Credential Can Be Sent to an Arbitrary Configurable Endpoint

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
Findings (13)

Tp4

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding

This mismatch is more security-relevant because the description frames the skill as a narrowly scoped beauty-video workflow, while the observed behavior reportedly includes broader generic text/image generation, account information access, wallet balance retrieval, model enumeration, and standalone media upload. Undisclosed expansion into account and upload operations can cause users or reviewers to underestimate data exposure, billing risk, and the effective permission surface.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

This mismatch is more security-relevant because the description frames the skill as a narrowly scoped beauty-video workflow, while the observed behavior reportedly includes broader generic text/image generation, account information access, wallet balance retrieval, model enumeration, and standalone media upload. Undisclosed expansion into account and upload operations can cause users or reviewers to underestimate data exposure, billing risk, and the effective permission surface.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

This mismatch is more security-relevant because the description frames the skill as a narrowly scoped beauty-video workflow, while the observed behavior reportedly includes broader generic text/image generation, account information access, wallet balance retrieval, model enumeration, and standalone media upload. Undisclosed expansion into account and upload operations can cause users or reviewers to underestimate data exposure, billing risk, and the effective permission surface.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
91% confidence
Finding

The skill advertises and demonstrates capabilities that require sensitive operations such as environment-variable access, filesystem access, shell execution, and network use, but it does not declare any explicit tool scope or allowed-tools boundary. That creates an over-privilege and auditability problem: an agent may invoke broader capabilities than reviewers or deployers expect, increasing the risk of secret exposure, unintended file modification, or unreviewed outbound requests.

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 hardcodes Chinese output and a specific production workflow without indicating that this is contingent on user preference. While not a direct code-execution issue, it can override user intent, reduce transparency, and make implicit activation more problematic by pushing users into a predefined language and output mode they did not choose.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill enables implicit invocation without any trigger scope, user-consent boundary, or contextual constraints. That can cause the agent to activate this workflow unexpectedly and steer requests into producing executable AI-HIVE commands or marketing-media outputs when the user did not explicitly ask for this skill, increasing the risk of unintended external API usage, cost-incurring actions, or misuse in sensitive content-generation contexts.

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 file is presented as a beauty-tutorial video skill, but the implementation bundles a broader general-purpose AI client with text chat, image generation, model enumeration, uploads, and other functions. This is dangerous because it expands the effective capability surface beyond the declared purpose, undermining least-privilege expectations and making it easier for downstream agents or users to invoke unrelated functionality under a narrowly branded skill.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

This code file contains natural-language descriptions and CLI prompts exclusively in Chinese, including the module docstring and later command help/output strings. The skill does not offer users an opt-in language selection or explain that it is intentionally limited to a Chinese-speaking audience, which can violate language/locale policy requirements.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The unrestricted chat command provides generic multimodal text-chat capability that is not constrained to beauty tutorials or short-video production. In context, this is more dangerous because the skill metadata markets a narrow creative workflow, so hidden general-purpose chat can be used to bypass policy scoping, perform unrelated tasks, or access broader model behavior than reviewers and users expect.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

The image-generation command exposes a generic image synthesis pipeline outside the stated beauty-tutorial video workflow. This broadens the skill into a more general media-generation tool, which can be abused for unrelated or policy-sensitive content creation while appearing to be an innocuous domain-specific skill.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

The user-facing description, argument parser description, skill name, and search aliases are all written in Chinese, with no indication that other languages are supported or that this locale restriction is optional. Under the stated policy, forcing a specific language without user opt-in is a natural-language policy concern.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill exposes a user-info endpoint that can retrieve account information and wallet balance, which is unrelated to the advertised beauty-video generation purpose. While not inherently malicious, this creates unnecessary access to potentially sensitive account metadata and increases the chance of unintended disclosure or misuse by an agent using this skill for a different pretext.

Content

No source excerpt is available for this finding.