ClawHub
Back to skill

Security audit

自动广告生成器

Security checks for vulnerabilities and agentic risk

Overview

This ad generator is broadly aligned with its purpose, but it needs review because its advertised local/offline PIL path can still try OpenAI/DALL-E before falling back locally.

Install only if you are comfortable with third-party image services receiving prompts or selected assets and with possible API credit usage. Treat the advertised PIL mode as not guaranteed offline unless the OpenAI/DALL-E attempt is removed or explicitly disabled, and review the Dreamina curl-to-bash installer before running it.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Description-Behavior Mismatch

Medium
Confidence
91% confidence
Finding
The script includes an undeclared OpenAI/DALL-E path and performs a direct network download of the generated image, which expands the skill's capabilities beyond the described local PIL or Dreamina workflow. This is dangerous because it can transmit prompt content and usage metadata to an external provider without clear disclosure, creating data-flow and supply-chain risk that users would not reasonably expect from the manifest.

Context-Inappropriate Capability

Medium
Confidence
88% confidence
Finding
The OpenAI integration is an unnecessary external capability relative to the stated scope, so the code grants the skill a broader data egress path than users and reviewers are led to expect. In a content-generation skill that may process product and marketing inputs, this mismatch increases the risk of silent third-party exposure and weakens trust boundaries.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The documentation encourages use of external APIs and credentialed services such as Dreamina and remove.bg without an explicit warning that user-supplied images and prompts may be transmitted to third parties. In a skill handling uploaded product or marketing assets, this can lead to unintended data disclosure and privacy/compliance issues.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The code sends generation prompts to OpenAI and then downloads the resulting image over the network without any user-facing warning or consent mechanism. In this skill context, users may assume background generation is local or uses only the named provider, so undisclosed outbound requests can expose sensitive campaign details, product information, or proprietary creative direction to third parties.

Static analysis

No suspicious patterns detected.