Context
GTA API was my first proper REST API project, built in the summer of 2021 to support a Discord bot I created called Klochon. The goal was to provide an easy way to retrieve GTA:SA files (specifically
gta3.imgfiles) to retrieve models straight from my Discord server.
Before this project, I had little experience with Express or Sequelize. I built this API from scratch to power one of Klochon's commands: /fetch. The bot could call the API to fetch game files to use for modding on Discord.
Why I built it
I was working on Klochon, a Discord bot for my own server. I had around a 500+ member community and getting these files quickly saved a lot of time for modders. The bot could fetch files directly from the API, making it easier for users to access game assets without needing to browse the .img files and extract the .dff and .txd manually.
I wanted to:
- Allow querying the database from Discord
- Serve original extracted game files directly to my users
This was my opportunity to learn more about:
- Building APIs
- Using Sequelize to model a MySQL DB
- Handling CORS so the bot could call the API
- Serving static files through Express
What I learned
Looking back, building gta-api taught me:
- How to structure a basic Express app
- How to model DB tables using Sequelize models
- How to serve static files
This was all before ChatGPT, a lot of it was figuring things out myself.
How it worked
After joining the Discord, users could use the fetch command to retrieve files. The bot would call the API and reply with the files.
Example:
/fetch gta3 file: Colt45
The bot would then query the API, which would return the files colt45.dff and colt45.txd from the gta3.img archive.
Wrapping up
Even though this was a small project, it was one of the first times I built something that connected a database, an API, and an actual live bot used by real people.
At the time I had little knowledge of API design or production systems! Everything was built from scratch through trial and error. This project taught me the basics of building services and made me a lot more comfortable with Express, Sequelize, and serving files through HTTP.
It also showed me the value of writing tools for communities I was part of. By automating something that used to be slow and manual (extracting files from gta3.img), the bot saved time for 500+ members and became a central part of my Discord server's modding workflow.
If I were to rebuild this today, I'd add pagination, proper validation, Swagger docs, rate limiting, support for all 3D era GTA games, and perhaps a web interface to browse these files.
Unfortunately, due to costs, I shut down the bot and since I never hosted the API, it no longer exists. Perhaps I'll revisit it in the future.