Support image viewing capability (view_image tool) like GitHub Copilot
Preflight Checklist
- [x] I have searched existing requests and this feature hasn't been requested yet
- [x] This is a single feature request (not multiple features)
Problem Statement
Problem
When using Claude Code in VS Code, users often need to analyze or describe images (screenshots, diagrams, UI mockups, etc.) during development. Currently, Claude Code lacks a built-in tool to read and interpret image files.
GitHub Copilot has a powerful built-in view_image tool that allows the assistant to read PNG/JPG/GIF/WebP images and describe their contents. This is extremely useful for:
- Analyzing UI/UX screenshots
- Reading diagrams and architecture drawings
- Debugging visual issues
- Understanding test output screenshots
Feature Request
Add a view_image (or similar) tool that Claude Code can invoke to read image files and provide their content/description to the language model.
Proposed Solution
Suggested Implementation
Following VS Code's Language Model Tools API (vscode.lm.registerTool), the tool could:
- Accept a file path as input
- Read the image file using
vscode.workspace.fs.readFile() - Convert the image to a format compatible with the underlying model (e.g., base64)
- Return the image data as a tool result
The tool declaration in package.json would use the languageModelTools contribution point.
References
- GitHub Copilot's
view_imagetool in VS Code provides this exact functionality - VS Code LM Tool API:
vscode.lm.registerTool()
Alternative Solutions
_No response_
Priority
High - Significant impact on productivity
Feature Category
CLI commands and flags
Use Case Example
_No response_
Additional Context
_No response_
This issue has 2 comments on GitHub. Read the full discussion on GitHub ↗