Back to skill

Security audit

小红书爆款种草视频生成与编辑

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly matches its video-generation purpose, but it needs review because it can send the user's API key to a configurable server and can write downloaded outputs using unsafe filenames.

Review this skill before installing. Use a virtual environment with pinned dependencies, keep the default AI Hive endpoint unless you fully trust a custom one, inspect `~/.ai-hive/config.json`, and prefer a dedicated empty output directory or `--no-download` until filename containment is fixed. Rotate the AI Hive API key if it may have been used with an untrusted base URL.

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 (3)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:87
Finding

Unpinned Third-Party Dependency Installation

Content
View full analysis
Remediation
View remediation
--hash=sha256: ``` 3. Install with hash verification: ```bash python3 -m pip install --require-hashes -r requirements.txt ``` 4. Pin and hash all transitive dependencies, not only `requests`. 5. Document an approved package index and avoid falling back to untrusted extra indexes. 6. Run dependency installation inside a dedicated virtual environment without administrative privileges. 7. Add automated dependency vulnerability and integrity scanning to the release process. ]]>

T09 · Insecure Skill Coding Practices

Error
Location
scripts/videogen.py:90
Finding

Bearer Credential Disclosure Through an Unrestricted Configurable API Endpoint

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
scripts/videogen.py:399
Finding

Output Directory Escape Through Unsanitized Remote Task Fields

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
  • 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)

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
94% confidence
Finding

This skill file contains a general text-chat pathway that can send arbitrary prompts and uploaded images to any TEXT model, which exceeds the declared Xiaohongshu video-generation purpose. Even if the current entrypoint does not expose it directly, embedding unrelated high-capability code in a skill increases the chance of accidental invocation, reuse by other components, or future exposure that turns a narrowly scoped media skill into a broader data-exfiltration or policy-bypass surface.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
92% confidence
Finding

The skill exposes operational commands that use shell, network access, file I/O, and environment-derived paths, but it does not declare any tool restrictions or allowed-tools scope. In an agent environment, this increases the blast radius of accidental or prompt-influenced execution because the agent may invoke broader capabilities than are actually needed for a user's request.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

The description is broad enough to match many ordinary requests about Xiaohongshu, RED, product videos, UGC, and ad creatives, which can cause the skill to be selected in situations where its code-capable workflow is unnecessary. Over-broad triggering is risky because it can route unrelated user prompts into a skill that encourages script execution and optional downloading, expanding exposure to unintended actions.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill name, description, headings, and examples are presented primarily in Chinese and oriented to 小红书 content, but the file does not state that the language is optional or limited to users who request Chinese output. This creates a natural-language policy risk because it appears to enforce a specific language/locale without explicit user opt-in or a documented locale restriction.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The file docstring states this script is a '通用 AI 能力调用工具' that wraps text chat, image generation, video generation, model querying, media upload, and task polling. The manifest, however, presents the skill as specifically for generating and editing Xiaohongshu/RED seeding and brand-cooperation videos, so this broader documented purpose does not match the skill’s declared scope.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The inline documentation explicitly frames the file as a general AI Hive tool for chat, image, and video operations via argparse subcommands. Later in the file, the generated skill configuration and skill-specific parser clearly implement a dedicated Xiaohongshu video skill, so the documentation actively describes a different intent than the file now serves.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The module docstring and user-facing instruction text are written exclusively in Chinese and present the tool as a general-purpose AI capability wrapper, but there is no indication that users may select another language or locale. For a broadly scoped skill, this amounts to a language/locale policy constraint expressed in natural language without opt-in or justification.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The user-info capability retrieves account information and wallet balance, which is not necessary for a Xiaohongshu video-generation/editing skill. Unneeded access to billing or account metadata violates least privilege and could expose sensitive operational information if an agent or downstream wrapper invokes it unexpectedly.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

Broad model enumeration lets callers inspect the provider's available models beyond the skill's intended fixed Xiaohongshu video use case. While not directly compromising secrets, it expands reconnaissance capability and undermines the assumption that the skill is narrowly constrained to specific reviewed models and workflows.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The image-generation command is unrelated to the stated role of generating/editing Xiaohongshu videos and broadens the skill into a multi-purpose content-generation client. In a security context, unnecessary capability expansion matters because it increases the available attack surface, allows use outside the reviewed purpose, and weakens least-privilege assumptions for agents selecting this skill.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

This markdown file contains user-facing natural-language content almost entirely in Chinese, with no indication that the skill is region-specific or that users can opt into another language. Under the policy rule for language or locale constraints, forcing a specific language without user choice can be a violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The example API key placeholder uses Chinese text ('请替换为你的APIKey'), which imposes a specific language in a user-facing configuration artifact. The file does not offer an alternative language option or explain a region-specific requirement, so it conflicts with the language/locale policy criteria.

Content

No source excerpt is available for this finding.