screen-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
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": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| list_windowsA | List all visible windows on the Hyprland desktop via hyprctl. Use the returned address with screenshot_window, or use "focused" to target the active window. |
| screenshot_regionA | Capture a rectangular screen region by pixel coordinates using grim (Wayland wlroots screencopy). Coordinates are in layout pixels. Returns the image inline as base64. |
| screenshot_windowA | Capture a specific window by title, class, address, or the keyword "focused". Uses hyprctl to resolve the window and grim -T to capture it. Returns the image inline as base64. |
| screenshot_fullscreenA | Capture the entire screen using grim. Returns the image inline as base64. |
| analyze_imageA | Send an image to the Gemini API for visual reasoning. Useful when the calling model has no native image encoder. Provide either a base64-encoded image or a local file path. |
| screenshot_and_analyzeA | Capture a screen region and immediately send it to Gemini for analysis in a single call. Combines screenshot_region and analyze_image. Useful for agents without native vision. |
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 6 tools
Each tool has a distinct purpose: listing windows, capturing by target (window/region/fullscreen), analyzing an existing image, and a combined capture+analyze convenience. The overlap between screenshot_and_analyze and screenshot_region+analyze_image is intentional but creates slight ambiguity for agents choosing between them.
All tool names follow a consistent snake_case verb_noun pattern (list_windows, screenshot_window, analyze_image). The compound screenshot_and_analyze is well-formed and consistent with the style, maintaining a clear and predictable naming convention.
With 6 tools, the server is well-scoped and each tool serves a clear, non-redundant function for screen capture and image analysis. The count is ideal for a focused utility server.
The tool surface covers the core lifecycle of capturing and analyzing screenshots, including window enumeration and multiple capture modes. Minor gaps exist, such as no explicit monitor listing or image format options, but the essential workflows are fully supported.