A bot for Jim's Garage Discord Server
TBD
The bot runs a FastAPI webserver in parallel to the bot, providing an API to interact with the bot.
The webserver defaults to port 8080, but you can customize this by setting
the API_PORT environment variable. That said, when running this as a
container, you should probably leave that alone and just map a different host
port if needed.
Automatically generated API documentation is available at:
/docs/redoc
The bot is packaged up inside a Docker image, which can be run via Docker, Docker Compose, or Kubernetes.
The following environment variables are supported:
| ENV VAR | Required | Use | Default |
|---|---|---|---|
| API_PORT | No | Set the port the API listens on (inside the container) | 8080 |
| BOT_TOKEN | YES | The token for your Discord bot | N/A |
| LOG_DIR | No | Directory where the bot's logs will be written | /app/log |
| LOG_FILE | No | Filename the logs will be written to | bot.log |
| LOG_LEVEL_FILE | No | Log level written to the log file | INFO |
| LOG_LEVEL_STDOUT | No | Log level written to stdout | INFO |
If all goes well, you should see logs like the following:
[2025-03-05 04:39:38] [INFO ] lib.bot: We have logged in as <your bot name shows up here>
[2025-03-05 04:39:39] [INFO ] uvicorn.access: GET /healthcheck HTTP/1.1 200 OKEach image will be published with four tags: latest, the major version, the
minor version, and the patch version. For example, for a version of 1.3.12,
the tags would be:
latestv1v1.3v1.3.12
This gives you flexibility in how discrete you want to be in choosing the
version of the bot that you run. I strongly discourage using latest. You
never know what might get written over that tag, and the next time you restart
your container, there would be a decent chance of pulling a broken or
unexpected image.
Make your own copy of the sample.env for yourself with your bot token.
The .env file is .gitignore'd to keep any tokens out of Git.
docker run --name bot \
--env-file .env \
-p 8080:8080 \
-dit \
ghcr.io/cyberops7/discord_bot:"${TAG}"The Compose file is set up to read the environment variables from a .env
file (to keep the bot token out of Git). A copy of the main sample.env file
is symlinked in the docker/ directory. Make your own .env file and
customize it all to your liking.
docker compose -f docker/docker-compose.yaml up -dA sample manifest is provided in kubernetes/. There are a few things that you'll need to customize:
- How you create your Kubernetes Secret for the bot token. The sample manifest leverages a 1Password operator to create a Secret directly from a 1Password vault.
- The storageClass of the PersistentVolumeClaim
I am using a .gitignore'd Kustomize manifest to add my correct 1Password
vault path. A sample Kustomize manifest is included if you are curious about
that.
To test the manifest generation using Kustomize:
kubectl kustomize kubernetes/To apply the manifest with Kustomize:
kubectl apply -k kubernetes/For details on contributing to the project, see CONTRIBUTING.
To see planned features and improvements, check out TODO.
Please submit Issues here in GitHub to track suggestions/bug fixes, etc. Once the initial backlog of ToDos is wrapped up, I plan to deprecate TODO.md and work off of GitHub Issues.