Security audit
Security checks for vulnerabilities and agentic risk
This is a coherent logo-design skill that discloses its use of image-generation tools and APIs, with no evidence of hidden persistence, deception, or unrelated data access.
Install this if you are comfortable using third-party image-generation services for logo work. Do not include confidential brand strategy, unreleased product names, or sensitive customer data in prompts unless you have approved the chosen provider and account credentials.
The skill requests Bash alongside Read/Write/Edit even though its primary purpose is design guidance. This expands the agent's capability to execute arbitrary shell commands from a design workflow, increasing the risk of command execution, filesystem changes, or chained abuse if prompts or future edits steer the agent beyond benign image-generation examples.
The trigger clause includes any design request containing broad keywords like "logo", "标志", "徽标", or "图标设计", which can overlap with ordinary conversation and unrelated design contexts. The file does not provide exclusion conditions or negative examples to clarify when the skill should not activate.
The skill instructs the agent to send prompts to third-party image-generation APIs using API keys, but it does not clearly warn users that their inputs and possibly sensitive branding concepts will be transmitted externally. In a tool-enabled agent context, this can cause unanticipated disclosure of proprietary or personal data to external services.
The skill includes a concrete external HTTP request using an Authorization header and user-supplied prompt content, which constitutes real data transmission to a third-party service. In context this is expected functionality for AI image generation, but it remains security-relevant because brand concepts, confidential prompts, and usage metadata may leave the local environment.
# 需先配置环境变量 DASHSCOPE_API_KEY
curl -X POST 'https://dashscope.aliyuncs.com/api/v1/services/aigc/text2image/image-synthesis' \
-H "Authorization: Bearer $DASHSCOPE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
The summary and description, along with the rest of the skill instructions, are presented only in Chinese, which effectively imposes a language constraint on users. The file does not state that the skill is Chinese-only, region-specific, or that alternative language support is available by user choice.
No suspicious patterns detected.