oJo MCP Server
Lets AI agents turn HTML into images, offering free live preview links and optional hosted PNG generation, reusable templates, and template management via the oJo API.
README
@ojodotso/mcp-server
A Model Context Protocol server that lets an
AI agent turn HTML into images. Live previews are free and need no account; an
oJo API key adds hosted PNG URLs and reusable templates through the user's own
account. Transport is local stdio; the API boundary is HTTP to
https://api.ojo.so/v1 only.
Quick start — no account needed
Point your MCP client at the package. With no API key the server starts in
preview-only mode: create_preview_link renders HTML into a live,
full-resolution image preview, free and unlimited, with no oJo account.
{
"mcpServers": {
"ojo": {
"command": "npx",
"args": ["-y", "@ojodotso/mcp-server"]
}
}
}
claude mcp add ojo -- npx -y @ojodotso/mcp-server
Adding a key
A key unlocks the permanent PNG URL — the thing you can put in an email, a
social card or an API response — plus reusable templates. Create one at
https://ojo.so/dashboard/api and pass it as OJO_API_KEY:
{
"mcpServers": {
"ojo": {
"command": "npx",
"args": ["-y", "@ojodotso/mcp-server"],
"env": { "OJO_API_KEY": "sk_..." }
}
}
}
claude mcp add ojo -e OJO_API_KEY=sk_... -- npx -y @ojodotso/mcp-server
Works the same for Claude Desktop, Cursor, and Claude Code.
Tools
Only create_preview_link is registered without an API key; the rest talk to
the oJo API, so advertising them keyless would offer tools that can only fail.
| Tool | Purpose | Credit | Needs a key |
|---|---|---|---|
create_preview_link |
Build a free, full-resolution /preview link for an HTML/Handlebars draft. |
— | no |
generate_image_from_html |
Render plain HTML (with inline CSS) to a PNG; returns its public URL. | 1 | yes |
generate_image_from_template |
Render an existing template, optionally overriding variables via modify. |
1 | yes |
list_templates |
List the templates available to the account. | — | yes |
get_template |
Fetch one template's decoded HTML and variable defaults. | — | yes |
create_template |
Create a reusable template from plain HTML/Handlebars. | — | yes |
The design loop
Iterate with create_preview_link, then render. It returns an
https://ojo.so/preview#draft=… URL that renders the draft live in the browser
through the same engine as the final PNG, at full resolution. The draft is
gzipped into the URL hash, so it never reaches a server and nothing is stored
— the link is self-contained and shareable with a human reviewer.
Both generate tools also accept inspect: true, which returns a 512px-wide
downscaled thumbnail alongside the URL. That is a confirmation that the render
succeeded, not a design surface — it is too small to judge layout or typography.
Guiding prompt
author_ojo_template — teaches the Handlebars helper catalog, the Template
Variables (schema/defaults) vs Modify Payload (per-render data) model,
viewport conventions, and the preview-first design loop.
Configuration
OJO_API_KEY(optional) — your oJo API key. Without it the server runs in preview-only mode; with it the full tool surface is registered.OJO_API_BASE_URL(optional) — defaults tohttps://api.ojo.so/v1.OJO_WEB_BASE_URL(optional) — the web app that serves/previewforcreate_preview_link. Defaults tohttps://ojo.so; point it at a localhttp://localhost:3000when developing against a self-hosted web app.
Develop
pnpm install
pnpm build
pnpm test
pnpm check
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.