Midjourney MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| COTLET_API_KEY | Yes | Your Cotlet.dev API key, obtained at https://cotlet.dev/app. Export it as COTLET_API_KEY. Without it, a tool call answers: "Set COTLET_API_KEY. Get a key at https://cotlet.dev/app". |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
Midjourney through Cotlet.dev. Call midjourney_guide first. Set COTLET_API_KEY (https://cotlet.dev/app).
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| midjourney_guideA | How to work in Midjourney (generate, vary, upscale, animate and organize images and videos): its screens and the sign-in link. Call this first. |
| midjourney_statusA | Check that your Cotlet.dev account and Midjourney are ready: account, credits and Midjourney connection. |
| midjourney_openA | Open a Midjourney page in your signed-in Cloud Browser (default https://midjourney.com) to work with generate, vary, upscale, animate and organize images and videos. |
| midjourney_readB | Read the current Midjourney page: its text and numbered buttons, links and fields. |
| midjourney_clickC | Click a numbered element on the current Midjourney page. |
| midjourney_typeC | Type text into a numbered field on the current Midjourney page. |
| midjourney_closeA | Close the Midjourney browser session when the work is done, so credits stop running. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 7 tools
Each tool has a clearly distinct role: guide, status, browser lifecycle, page reading, and element interaction. Open/read/click/type are standard browser primitives with clear boundaries and no overlap.
All tools use a consistent midjourney_ prefix followed by a simple action verb. The pattern is predictable and easy to scan.
Seven tools is well-scoped for a browser-automation wrapper around Midjourney. Setup, session management, and page interaction each have an appropriate dedicated tool.
The core browser session and interaction loop is covered, but important Midjourney workflows are missing: no screenshot or visual inspection for generated images, no wait/poll tool for long generations, no upload for reference images, and no scroll/navigation helpers. Agents can work around some gaps with repeated reads, but not all.