使用Mistral Document AI从图片中提取文字（OCR）
AI agents call extract_text_from_image to retrieve information from Article Quadrant Analyzer MCP Server without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool performs optical character recognition to extract and retrieve text content from images. It is a read-only operation with no capability to modify state, execute external commands, delete data, or trigger financial transactions.
From the tool's definition Tool name and description indicate OCR functionality: 'extract_text_from_image' and '使用Mistral Document AI从图片中提取文字（OCR）' (extract text from images using Mistral Document AI OCR).
Attacks that exploit this kind of access
使用Mistral Document AI从图片中提取文字（OCR）. It is categorised as a Read tool in the Article Quadrant Analyzer MCP Server MCP Server, which means it retrieves data without modifying state.
Register the Article Quadrant Analyzer MCP Server MCP server in PolicyLayer and add a rule for extract_text_from_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 Article Quadrant Analyzer MCP Server. Nothing to install.
extract_text_from_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 extract_text_from_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 extract_text_from_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.
extract_text_from_image is provided by the Article Quadrant Analyzer MCP Server MCP server (vincent623/-article-quadrant-analyzer-mcp-server). 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 →