ComfyUI MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| COMFYUI_CONFIG | Yes | Path to the ComfyUI configuration file (config.json) |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Server capabilities have not been inspected yet.
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| comfy_submit_workflowB | Submit a complete workflow JSON to ComfyUI for execution. Supports parameter overrides for dynamic modifications without editing the workflow structure. |
| comfy_generate_simpleC | Quick image generation using pre-configured workflow templates (flux_txt2img, sd15_txt2img, sdxl_txt2img, basic_img2img). Ideal for common use cases without managing workflow JSON. |
| comfy_get_statusA | Get the current status and progress of a specific generation or the overall queue. Returns queue position, progress, and output paths when available. |
| comfy_wait_for_completionA | Block until a generation completes or fails. Returns final outputs with image paths. Useful for synchronous workflows. |
| comfy_list_modelsB | List available models, checkpoints, LoRAs, VAEs, and other resources in the ComfyUI models directory. Supports filtering by type and name. |
| comfy_save_workflowC | Save a workflow JSON to the MCP library for later reuse. Includes metadata like description and tags for organization. |
| comfy_load_workflowB | Load a saved workflow from the MCP library by name. Returns the workflow JSON and metadata. |
| comfy_list_workflowsA | List all saved workflows in the MCP library. Supports filtering by name, description, or tags. |
| comfy_delete_workflowA | Delete a saved workflow from the MCP library. Requires confirmation for safety. |
| comfy_get_queueA | Get detailed information about the current generation queue, including running and pending items. |
| comfy_cancel_generationB | Cancel a specific generation or interrupt the currently executing generation. Can optionally remove from queue. |
| comfy_clear_queueA | Clear all pending items from the queue (does not affect currently running generation). Requires confirmation. |
| comfy_upload_imageC | Upload an image to ComfyUI's input folder for use in workflows. Supports custom filenames and overwrite control. |
| comfy_get_output_imagesA | List recent output images from ComfyUI's output folder. Returns full Windows paths that Claude Desktop can read. |
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 14 tools
Each tool has a clearly distinct purpose with no ambiguity. For example, comfy_generate_simple is for quick image generation, comfy_submit_workflow handles complete workflow execution, and comfy_get_status monitors progress, all targeting different aspects of the ComfyUI workflow lifecycle. The descriptions clearly differentiate between queue management, workflow management, generation, and output handling.
All tools follow a consistent comfy_verb_noun pattern with snake_case throughout. The naming convention is predictable and readable, such as comfy_list_workflows, comfy_save_workflow, and comfy_cancel_generation. This uniformity helps agents easily identify and select the appropriate tools.
With 14 tools, the count is well-scoped for managing ComfyUI workflows, covering generation, queue management, workflow storage, and output retrieval. Each tool earns its place by addressing specific needs in the image generation pipeline, from submission to completion, without being excessive or insufficient.
The tool surface provides complete coverage for the ComfyUI domain, including CRUD operations for workflows (list, load, save, delete), generation lifecycle (submit, cancel, wait, status), queue management (get, clear), and output handling (get images, upload image). There are no obvious gaps, enabling agents to handle end-to-end workflows seamlessly.