dsh-media-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| MEDIA_GEN_CONFIG | No | Path to the config file. Overrides the default location. | |
| MEDIA_GEN_API_KEY | No | Required API key. Never printed by any tool. | |
| MEDIA_GEN_BASE_URL | No | Overrides the preset's API URL. | |
| MEDIA_GEN_PROVIDER | No | Which provider preset to use. Default `ark`. | ark |
| MEDIA_GEN_IMAGE_SIZE | No | Default image size, in WxH format. | |
| MEDIA_GEN_OUTPUT_DIR | No | Overrides both default output directories for images and videos. `~` is expanded. | |
| MEDIA_GEN_IMAGE_MODEL | No | Default model for image generation. | |
| MEDIA_GEN_VIDEO_MODEL | No | Default model for video generation. | |
| MEDIA_GEN_IMAGE_TIMEOUT_MS | No | Milliseconds one image call may take. Default `180000`. | 180000 |
| MEDIA_GEN_VIDEO_TIMEOUT_MS | No | Milliseconds the whole video flow may take. Default `720000`. | 720000 |
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
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 |
|---|---|
| generate_imageA | Generate an image from a text prompt and save it to disk. Call this when the user asks to draw, create, render, illustrate or generate a picture, photo or artwork. Returns the absolute paths of the saved files; pass one to read_image to view it. Each call costs real money, so call it once per explicit user request. |
| generate_videoA | Generate a short video from a text prompt and save it to disk. Asynchronous: the provider renders in the background and this call waits, typically one to three minutes. Call it only when the user explicitly asks for a video — reference images and videos are not supported. Returns the absolute path of the saved file, the task id and how long it waited. |
| get_configA | Report the current configuration: provider, base URL, the image and video models in use, their defaults, the output directory, whether an API key is present (never its value), and the catalogue of known models with their constraints. Call this before changing settings, or when the user asks which model is in use. |
| set_configA | Change settings and persist them to the config file. Call this when the user asks to switch a model, change a default size, output directory or timeout, or change provider. Only the fields passed are changed and everything else is preserved. Changes take effect on the next call with no restart. |
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 4 tools
Each tool targets a distinct operation: image generation, video generation, config retrieval, and config modification. No overlap exists, making selection unambiguous.
All tools follow the verb_noun pattern: generate_image, generate_video, get_config, set_config. The naming is consistent and predictable.
With 4 tools, the server is well-scoped for its purpose—media generation plus configuration management. Each tool is essential and there are no redundant or missing core operations.
The core generation and configuration workflows are covered, but the reference to a read_image tool in generate_image's description suggests a potential gap for viewing outputs. No explicit file listing or deletion tools are present, though these may be outside the server's intended scope.