Skip to main content
Glama
Sign In
shotbox

shotbox

by mishan
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Local

Related Servers

Alternatives to shotbox

  • linux-x11-sandbox

    Wangxiaoxiaoa
    A
    license
    Not graded
    quality
    A
    maintenance
    Enables AI agents and automation tools to interact with GUI applications in isolated X11 displays via MCP, supporting input, window management, screenshots, clipboard, and accessibility actions.
    4
    MIT

  • kwin-mcp

    isac322
    A
    license
    Not graded
    quality
    A
    maintenance
    Enables AI agents to automate Linux desktop GUI by launching and interacting with Wayland applications in isolated virtual KWin sessions, or connecting to live desktops for collaborative automation.
    155 PyPI
    77
    MIT

  • wayland-computer-use-mcp

    Niklas-Ehrenfried
    A
    license
    Not graded
    quality
    A
    maintenance
    Enables AI agents to inspect, automate, and test desktop GUI applications on Wayland through window/desktop streaming, semantic AT-SPI accessibility tree inspection, and boundary-clamped mouse, keyboard, drag, and scroll input injection. It also manages application process lifecycle, captures window frames, and registers Python GUI scripts as native Linux desktop launchers.
    1
    MIT

  • deskwright

    tristanmuzzu
    A
    license
    A
    quality
    B
    maintenance
    Enables AI agents to operate a real GNOME Wayland desktop through accessibility-tree widget actions, pointer/keyboard input, OCR, window management, and screen capture, optionally on a private headless session.
    33
    21 PyPI
    11
    Apache 2.0

Related Servers

  • deskcheck

    alexwboles
    A
    license
    Not graded
    quality
    C
    maintenance
    Enables coding agents to run deterministic desktop GUI tests over MCP: launch allowlisted apps, execute scripted click/type/key/screenshot/assert steps, and receive pass/fail reports for their build-fix loop.
    MIT

  • mango-mcp

    Anti-2k
    A
    license
    Not graded
    quality
    B
    maintenance
    Enables AI agents to use an isolated private headless Wayland desktop where they can launch apps, take screenshots, manage windows, and drive mouse and keyboard without affecting the user's own session.
    MIT

  • uisight

    mTakeda89733
    A
    license
    Not graded
    quality
    A
    maintenance
    Runs live mobile and desktop browser sessions that measure a web UI — contrast ratios, touch-target sizes, overflow and theme drift — and hands the results back to an agent as precise text with selectors. A human and the agent watch the same session, so pinned notes and the exact frame at that moment are readable too.
    268 npm
    MIT

  • blindqa

    DevZonayed
    A
    license
    Not graded
    quality
    A
    maintenance
    Enables AI coding agents to drive a real browser like a person — crawling every page and role, running real user journeys, and auditing controls a DOM-based test misses — so UI, accessibility and permission bugs are found and only the screens affected by recent changes are re-tested. Runs are script-driven and read-only by default, so testing costs no model tokens and nothing about the app is sent to an AI.
    MIT
Deploy Server
One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

TDQS

A3.9/5.0

Scored across 18 tools

Disambiguation4/5

Most tools target clearly distinct actions (start_session/stop_session, click_element, scroll_wheel, press_keys, type_text, record_video, record_audio, the three check_* linters), and the descriptions explicitly draw boundaries (e.g. type_text vs press_keys, screenshot vs accessible_tree). The main overlap is wait_for, which reimplements waiting on announcements, log lines, focus, and screen changes that other tools also observe, creating some potential misselection.

Naming Consistency5/5

Every tool uses snake_case with a predictable verb_noun or verb_object pattern: start_session, stop_session, spawn_program, get_program_output, get_accessible_tree, take_screenshot, click_element, scroll_wheel, press_keys, record_video, record_audio, use_gamepad, check_accessible_names, check_contrast, check_focus_order. wait_for is the only slight deviation but remains readable and consistent in style.

Tool Count4/5

18 tools is on the heavier side but each earns its place across the distinct sub-domains of session management, program spawning, input simulation, observation, recording, and accessibility checks. It is slightly large but not padded; nothing appears redundant enough to remove.

Completeness5/5

The surface covers the full lifecycle: session start/stop, program spawn/output, input (text, keys, clicks, scroll, gamepad), observation (accessible tree, screenshots, video, audio, announcements), flexible waits, and three accessibility linters. Coverage is comprehensive for a GUI/accessibility testing domain with no obvious dead ends.

Maintenance

ActivityActive
ResponsivenessNo recent community issues

Resources