Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    godot ai playtest

    1

    Automated Godot 4 playtesting bot that audits UI, finds translation overflows, and verifies level flow.

    Free

    Security scanned
    godot ai playtest

    godot ai playtest

    Example session with this skill installed

    Run a playtest on the current project and tell me if the new inventory screen broke the French translation layout.

    • Read your context and instructions
    • Compiled the godot ai playtest

    I have executed the playtest suite for the French (fr) locale. The bot successfully navigated through the main menu to the inventory screen.

    Playtest Report: French Locale (fr)

    SceneStatusResult
    main_menu.tscnPASSFinished in 1.2s
    inventory.tscnFAILoverflow detected
    level_01.tscnSKIPPEDFlow interrupted

    UI Issues

    [inventory.tscn] overflow: The label "Equipement de protection individuelle" is cut off by the ItemSlot container.

    • [inventory.tscn] offscreen: The "Retour au menu principal" button extends 12px past the right screen edge.
    • [inventory.tscn] raw_key: Found KEY_REPAIR instead of a translated string in the tooltip.

    Sequence of play

    1. Clicked "Jouer"
    2. Clicked "Inventaire"
    3. Found overflow on item description.
    4. Attempted to click "Retour", but button was partially off-screen.

    Next steps

    1. Widen the DescriptionBox in inventory.tscn or enable text wrapping for the French locale.
    2. Update fr.translation to include the missing KEY_REPAIR value.
    3. Re-run tools/playtest.sh to verify the fix.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Testing Godot 4 games manually is slow, prone to human error, and impossible to scale across multiple languages. You cannot see the UI you are building, leading to off-screen text, overlapping labels, and broken level flows that only surface after a build. This skill installs an automated playtesting bot that "sees" your UI and interacts with your game like a real player to catch regressions immediately.

    What it does

    • Automated Input simulates real key presses, clicks, and taps to navigate scenes and finish levels.
    • UI Auditing detects off-screen text, overlapping labels, overflow cut-offs, and untranslated raw keys.
    • Multi-language Testing runs every scene through every defined locale to catch translation-specific layout breaks.
    • Visual Validation captures screenshots automatically when new text appears for manual review.
    • Fuzz Testing applies random inputs to discover crashes or stuck states in the game logic.

    How it works

    1. Initialize the project by installing the playtest runner, config files, and GitHub Action workflow.
    2. Define Scenes in the configuration file to set the order of play from main menu to final level.
    3. Author Brains for each scene, teaching the bot how to play (e.g., click specific labels or press keys).
    4. Execute the playtest.sh script to run the bot headlessly and generate a Markdown report and screenshots.

    Frameworks & tools

    Built for Godot 4 (4.3+) GDScript projects. Integrates with GitHub Actions for CI/CD and works with standard Godot Control nodes for UI inspection.

    Why this beats prompting it yourself

    Writing a custom playtest runner from scratch requires complex input simulation and UI tree traversal logic. This skill provides a tested architecture that catches specific Godot UI bugs, like MENU_START raw key leaks, that standard prompts would miss.

    Use cases

    • Smoke testing a new level to ensure it is actually beatable.
    • Auditing UI layouts across five different languages for text overflow.
    • Verifying that a UI refactor didn't break the main menu navigation.
    • Running automated regression tests on every GitHub pull request.

    Known limitations

    Does not support Godot 3. Games using _draw() for UI require manual probes. The bot cannot judge game "feel" or difficulty balance.

    Godot 4 only. UI must use Control nodes for automatic detection; _draw() requires custom probes. Requires Godot binary in PATH or specified location. Headless execution only.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Free to use

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed today

    • Free with an Agensi account

    Accounts
    GitHub

    Permissions

    Terminal / Shell
    Network Access
    Read Files
    Write Files

    Allowed Hosts

    github.com
    Listedtoday

    What's inside

    Frequently Asked Questions

    This skill is purpose-built for Godot 4.3 and newer GDScript projects. It relies on the modern Godot Control node system to inspect UI trees and simulate inputs, so it is not compatible with Godot 3.x or projects using custom _draw() calls for UI.

    While the bot uses 'brains' to navigate levels, it is designed to catch objective bugs like UI overflows, crashes, and broken level transitions. It cannot provide feedback on subjective qualities like game balance, difficulty, or 'fun' factor.

    The skill includes the core playtest runner, a set of template 'brains' for common game scenes, a local execution script, and a pre-configured GitHub Action workflow for CI/CD integration. You also receive a documentation guide for configuring your specific scene order.

    Setup involves running an initialization script that adds the runner to your project, then defining your scene flow in a configuration file. For complex levels, you will need to 'author' basic logic to tell the bot which UI elements to interact with or which keys to press to reach the goal.

    Yes, one of the core features is multi-language auditing. The bot can automatically cycle through all your defined locales and capture screenshots of every UI state, allowing you to instantly spot text that overflows its containers in languages like German or localized strings that failed to load.

    New in Game Development

    Free