分析截图并提取 UI 组件为可复用的 Tailwind HTML
AI agents call analyze_image to retrieve information from Componentize without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool reads and analyzes design screenshots to identify and extract UI components, returning structured data (Tailwind HTML). It performs no modifications to the source images, does not execute arbitrary code, and does not create persistent changes to external systems. This is a retrieval and analysis operation with no blast radius beyond providing information.
From the tool's definition Tool description states 'analyze' and 'extract' UI components from screenshots, with no modification or deletion of source data. The sibling tool 'export_components' suggests this tool performs analysis/extraction only.
Attacks that exploit this kind of access
分析截图并提取 UI 组件为可复用的 Tailwind HTML. It is categorised as a Read tool in the Componentize MCP Server, which means it retrieves data without modifying state.
Register the Componentize MCP server in PolicyLayer and add a rule for analyze_image: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Componentize. Nothing to install.
analyze_image is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the analyze_image rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for analyze_image. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
analyze_image is provided by the Componentize MCP server (monlohua/componentize). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →