Image Viewer
Show screenshot and design files inline in agent chat by passing a local file path instead of manual uploads.
Overview
InChat Image Viewer MCP is a MCP server for the Build phase that renders local image files inline in AI chat from a file path.
What is this MCP server?
- InChat Image Viewer MCP v1.0.1 on npm as inchat-image-viewer-mcp
- Inline image rendering in AI chat from a simple filesystem path
- stdio MCP transport for local agent setups
- No API key in server.json—path-based viewing only
- Server version 1.0.1
- npm identifier inchat-image-viewer-mcp
- stdio transport
What problem does it solve?
Agents cannot see your screenshots or UI mocks unless you awkwardly upload or describe pixels by hand.
Who is it for?
Solo builders iterating on UI, debugging visual bugs, or sharing local mockups with an MCP agent.
Skip if: Remote URL galleries, automated image editing pipelines, or teams that need cloud asset management.
What do I get? / Deliverables
You pass a file path and the chat shows the image so review and iteration stay in one thread.
- Inline image display inside agent conversations
- Faster visual review loops without leaving the chat
Recommended MCP Servers
Journey fit
Visual verification while implementing UI and assets is core to Build; agents otherwise cannot see what you are referencing on disk. frontend fits best because the primary win is reviewing layouts, mocks, and exported images during implementation.
How it compares
Local path image preview for chat, not a DAM or browser automation skill.
Common Questions / FAQ
Who is io.github.OguntolaIbrahim/image-viewer for?
Developers using MCP clients who want agents to see local screenshots and design files without manual image uploads.
When should I use io.github.OguntolaIbrahim/image-viewer?
Use it during frontend Build work when you need the agent to inspect a specific image file on disk by path.
How do I add io.github.OguntolaIbrahim/image-viewer to my agent?
Add inchat-image-viewer-mcp from npm as a stdio MCP server in your client config, then invoke tools with valid local image paths.