Skip to content

Educational application aimed at automating user-defined workflows for the mobile game, "Granblue Fantasy", using a variety of CV technologies in the backend such as OpenCV, PyAutoGUI and EasyOCR and a frontend coded in Typescript.

Notifications You must be signed in to change notification settings

steve1316/granblue-automation-pyautogui

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Granblue Automation using Template Matching (It is like Full Auto, but with Full Customization!)

GitHub commit activity GitHub last commit GitHub issues GitHub pull requests GitHub

Discord here: https://discord.gg/5Yv4kqjAbm

Android version here: https://github.com/steve1316/granblue-automation-android

Checkout Granblue Automation Statistics and its project over at https://github.com/steve1316/granblue-automation-statistics

uClriDn77l.mp4

This Python application is designed for educational research purposes on studying how to automate certain workflows via image template matching using OpenCV. The application uses PyAutoGUI for direct mouse control, hence why it is recommended to run this on a separate machine than the one that you use daily. This can be circumvented by running this on a virtual machine like VMWare Workstation Player so you can keep using your main computer without interruption.

Screenshots are taken and cropped by PyAutoGUI for OpenCV to perform image template matching. This will determine where the bot is currently at and will inform the bot on what to do next from there.

There is a feature already in-game that can automate gameplay called "Semi/Full Auto" but does not offer any way to customize what each character does on a turn-by-turn basis. This program's primary goal is to provide that customization. Users can create their own combat scripts using predefined case-insensitive keywords and can indicate which turns the bot will execute their script, somewhat akin to constructing pseudocode.

For example:

// This is a comment. The bot will ignore this line.
# This is also a comment.

Turn 1:
    // On Turn 1, the following commands will be executed in order:
    // 6th Summon is invoked, character 1 uses Skill 2 and then Skill 4,
    // and finally character 3 uses Skill 3.
    summon(6)
    character1.useSkill(2).useSkill(4)
    character3.useSkill(3)
end

# The bot will keep clicking the Attack button until it reaches the 5th turn.
Turn 5:
    character2.useSkill(2)
end

// Use the exit keyword to leave the raid without retreating.
// Useful when you want to farm multiple raids at once.
exit

Disclaimer

By downloading this program, you consent to your account potentially getting flagged for excessive amounts of farming for multiple hours straight and banned in the next banwave by KMR. I hold no responsibility for how much or how long you use this program for. I trust you have the self-control necessary to only farm in reasonable bursts of time with breaks in between and to always be alert for when the bot encounters the CAPTCHA.

Table of Contents

Features

  • Customize what skills to use during each turn in a user-created plan. Users can select which plan to use when starting the bot.
  • A launchable GUI to keep track of logs, adjust settings and selecting what combat script to use.
  • Farm user-defined amounts of specified materials from the supported Farming Modes.
  • A user-defined timer for how long the bot should run for.
  • Support for the following game modes:
    • Quest
    • Special
    • Coop
    • Raid
    • Event
    • Guild Wars
    • Rise of the Beasts
    • Dread Barrage
    • Proving Grounds
    • Arcarum
    • Replicard Sandbox Part 1
    • Replicard Sandbox Part 2
  • Alert for when anti-bot CAPTCHA pops up.
  • Discord integration for informing you of status updates like loot drops via private DMs.

Wiki

Visit the Wiki here for detailed documentation and examples.

Requirements

Visit the Installation Instructions page here.

Instructions

  1. Download the latest release by heading to the Releases page and downloading the latest .zip file. You will know its the correct one to download as it has a .exe executable file in it called Granblue Automation.exe. Alternatively if you want to build it by yourself, download the entire project itself and then head over to Build Instructions and follow the steps. Executing yarn build will create the .exe and the necessary files in /src-tauri/target/release/.

  2. Make sure you installed the project dependencies by having Python 3.8.3 installed and ran pip install -r requirements.txt as stated above.

  3. Open up the game on a Chromium-based browser and log in if you haven't already done so. Click away any daily log in popups until you are at the Main/Home screen.

    1. [REQUIRED] Make sure the window size is set to this or else the bot will not be able to detect the "Home" button and proceed any further:

      Correct Window Size

    2. [REQUIRED] Make sure that BOTH of the Auto Restore settings are enabled in the ingame settings.

      Auto Restore Settings

  4. Now open up the program executable and follow the onscreen instructions and checking that the Settings page is filled out. You will be informed that the bot is ready to start by the status message at the top of the window.

    1. (Optional) Additionally, you can check the Extra Settings page for more general settings like Twitter, Discord and Configuration settings.
  5. You can now head back to the Home page of the program and hit the "Start" button to begin.

    1. Message logs are stored in the /logs/ folder after the bot ends or encounters an error/unexpected situation.

Instructions for farming Raids (optional, only if you want to farm Raids)


Instructions to set up Discord integration


How to create my own Combat Script?


What Missions/Items/Summons are supported?


Virtual Machine Setup (optional)

Build Instructions

Note: If your changes are only in the Python files, you can just replace the files in your GA folder. But if the Javascript files were changed, then a build is needed.

Note 2: Adjust the /src-tauri/tauri.conf.json for build-specific adjustments to your environment.

Note 3: Adjust /src-tauri/tauri.conf.json and /src-tauri/update.json to update version number of your build. You would also need to sign your update as of Tauri v1.0 with a private key generated by you.

Note 4: Tauri does not support cross-platform building. This means when you build on a Windows platform, you get a Windows release. Same goes for Linux and MacOS. GitHub actions as of now are using the windows-latest environment to build a Windows release. If you need to build for Linux / MacOS, you would need to adjust how the frontend calls the main script, sets up the /logs/ folder, and kills the process using its PID as right now it is using Windows-specific commands to do so (Powershell) in StartHelper.tsx and ExtraSettings.tsx.

  1. Open up root folder in VSCode. Type yarn install in terminal to install Javascript dependencies.
  2. Then follow your environment's instructions for Tauri.
  3. Now refer to the following 2 commands below for your use case.
  • yarn start Starts up the Metro server for hot code reloading.

  • yarn build Bundles and builds the executable in /src-tauri/target/release/. Also includes a installer in ../bundle/ as well.

Technologies Used

  1. Python - The main language
  2. Qt - Application development framework for the GUI (old frontend)
  3. Typescript - Language for the new GUI (new frontend)
  4. Tauri - Toolkit to transform the Typescript frontend framework into a desktop application
  5. PyAutoGUI - Primarily for screenshot capturing and mouse control
  6. pyclick - For making mouse movements human-like via Bezier Curves
  7. OpenCV-Python - Image template matching
  8. EasyOCR - For text recognition and detection
  9. Twitter Standard API 1.1 - For searching and parsing texts for Raid room codes to join
  10. VMWare Workstation Player - For virtualizing the program to circumvent control of the main cursor

About

Educational application aimed at automating user-defined workflows for the mobile game, "Granblue Fantasy", using a variety of CV technologies in the backend such as OpenCV, PyAutoGUI and EasyOCR and a frontend coded in Typescript.

Topics

Resources

Stars

Watchers

Forks

Packages

No packages published