Command
A command that asks a vision tool to analyse a local image using a natural-language question.
Command
A command that asks a vision tool to analyse a local image using a natural-language question.
Command
A command for analysing charts and other data visualisations, including trends, unusual values, and business-relevant findings.
Command
A command for interpreting technical diagrams such as architecture drawings, flowcharts, UML diagrams, and database relationship diagrams.
Command
A command for comparing two user-interface screenshots and listing their visual differences by severity.
Command
A command for diagnosing error screenshots with a vision MCP server. It examines stack traces, console output, or error dialogs in an image and suggests fixes.
Command
A command that extracts all visible text from a screenshot while preserving its original layout and structure. OCR, or optical character recognition, is the process of reading text in images.
Command
A command that turns a user-interface screenshot into code or a written design description. A user interface is the visible part of an application that people interact with.
MCP server Claude CodeCodexCursor +2
MCP server exposing 7 vision tools (analysis, UI→code, OCR, error diagnosis, diagram, data-viz, diff) — works with any Anthropic-Messages-compatible vision model. Runs locally from the @spock-wen/vision-mcp-server npm package.