About Imejis
Imejis is an MCP server for AI agents to design and render branded images, aimed at teams that need consistent visual assets. It supports template-based and bulk rendering with fixed layouts, plus version management and share links.
Key features
- 01Discover components and create designs
- 02Create designs from templates and edit them
- 03Render single images or batches
- 04Manage design versions and rollbacks
- 05Manage assets, fonts, and templates
Use cases
- 01Create branded social images
- 02Generate images for multiple data variants
- 03Make design variations in different sizes
- 04Share a public design page
MCP support
It provides a hosted Streamable HTTP MCP server with OAuth 2.0 for client connections. Clients can use tools for design, rendering, publishing, asset management, and account information.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://api.imejis.io/api/mcpclaude mcp add --transport http io-imejis-imejis https://api.imejis.io/api/mcp{"mcpServers":{"io-imejis-imejis":{"type":"http","url":"https://api.imejis.io/api/mcp"}}}Open in Cursor{"servers":{"io-imejis-imejis":{"type":"http","url":"https://api.imejis.io/api/mcp"}}}Open in VS Code{"mcpServers":{"io-imejis-imejis":{"type":"http","url":"https://api.imejis.io/api/mcp"}}}{"mcpServers":{"io-imejis-imejis":{"serverUrl":"https://api.imejis.io/api/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
Details
From the project READMEImejis MCP Server
Give your AI agent a design studio. The Imejis MCP server lets Claude, ChatGPT, Cursor, or any MCP client design and render real, on-brand images for you.
Your agent designs a template once, then renders it in bulk with the same layout every time, only the data changing. Unlike a plain image model, the output is deterministic and pixel-perfect, so your brand never drifts.
- Endpoint:
https://api.imejis.io/api/mcp - Transport: Streamable HTTP
- Auth: OAuth 2.0 (RFC 8707 audience binding, RFC 9728 protected-resource metadata)
- Hosted: yes, nothing to install
- Landing page & quick tour: https://www.imejis.io/agents
- Server card: https://www.imejis.io/.well-known/mcp/server-card.json
This is a hosted MCP server. This repository holds its public metadata (
server.json) and connection docs; the server runs atapi.imejis.io.
Connect
Add the endpoint as a remote MCP server and authorize once over OAuth.
- Claude: Settings → Connectors → Add custom connector → paste
https://api.imejis.io/api/mcp→ approve the sign-in. Full guide - ChatGPT: Settings → Connectors → add an MCP server → enter the URL → sign in. Full guide
- Cursor: Settings → MCP → Add new server → paste the URL → approve. Full guide
What your agent can do
Once connected, the agent discovers the full catalog, designs, and renders:
- Discover:
get_design_guide,list_component_types(all 27 components: text, images, QR codes, barcodes, tables, and 12+ chart types), generated from the same registry the renderer validates against, so it never drifts. - Design:
create_design,create_design_from_template,update_design,duplicate_design,preview_design, pluscreate_design_variationto add another size (it reflows the master's layout onto the new canvas). - Render:
export_design(one image),get_render_url(a permanent URL that re-renders per request with field overrides),get_stored_image_url(stores the render, returns a hosted or expiring URL) andbulk_export(up to 25 renders in one call — several sizes, several data variants, several designs, or the ordered slides of a carousel). - Publish: every edit publishes a numbered version you can name, list,
read back and roll back (
publish_design,list_design_versions,get_design_version,rollback_design).create_share_linkgives the design a public page anyone with the URL can open, andcreate_template_from_designmakes it reusable across the workspace. Both serve a pinned version, soupdate_share_link/sync_template/set_live_versioncontrol exactly what the public sees. - Gallery: ask an admin to list a template publicly with
submit_template_for_reviewand follow it withlist_template_reviews. Approving is an admin action in the imejis dashboard — there is no tool for it, so an agent can ask but never publish to the gallery itself. - Assets & account: manage files, fonts, templates, API keys, and read plan, credits, and quota.
Anywhere a tool takes a design id, both ids work: the one a list returns or the short code from a render URL.
It also ships ready-made prompts: create_social_post, make_og_image,
edit_design, turn_design_into_api and export_design_image.
Try these
- "Create a 1200×630 OG image for a post titled 'How MCP works' and export the PNG."
- "Make a QR code that links to imejis.io/agents and give me the SVG."
- "Build a bar chart from this data, then render one image per row of my CSV with the same layout."
Prefer code?
The same renderer has a REST API with tested examples in cURL, Node.js, Python, Ruby, Go, PHP, Java, and C#: https://www.imejis.io/apis
Pricing
The free plan includes 100 renders per month, no credit card. Agent renders count against your plan the same as a direct API call. See https://www.imejis.io/pricing.
Links
- Website: https://www.imejis.io
- AI Agents: https://www.imejis.io/agents
- API docs: https://www.imejis.io/apis
- Components: https://www.imejis.io/components
- Privacy: https://www.imejis.io/privacy-policy
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.