Bluesky Python Bot

This repository contains a starter template for creating a Python bot on Bluesky.

💡 Some possible bots:

• Listens to the firehose and replies to every post that contains "tldr-bot" with a summary of the thread it was tagged in (see examples/firehose-bot.py)
• Posts a random image every hour (see examples/cron-bot.py)
• Listens to the firehose to see what topics are trending, posts a list of trending topics every 5 minutes

Set Up

1. Create a new Bluesky account for your bot on the app or website (https://bsky.app)
2. Generate an App Password for your bot account in settings (this is just to protect your real password)
3. Make a copy of the example .env file by running: cp .env.example .env
4. Set your bot's username and app password in .env
5. Create a new virtual environment with venv or your favorite package manager.

Using venv: python3 -m venv .venv source .venv/bin/activate pip install -r requirements.txt When you're done: deactivate

Building your bot

Modify the main.py script however you like to make this bot your own. The /examples director has firehose-bot.py and cron-bot.py - if you want your bot to respond to some kind of event live, you want firehose-bot. If instead you want to do something on a timer, use cron-bot. You might also want to do something that's kind of both (i.e. it listens to the websocket to index/aggregate data and then does something on a schedule, like say what the sentiment is for a given search term)

Running your bot

You can run the script locally: python3 main.py. You should see a smiley emoji posted to your Bluesky account.

Data

Data availability is a huge perk of atproto--data can be accessed by crawling the repos yourself, ingesting the firehose, querying bluesky's AppView, or querying someone else's service (like a labeler or feed generator).

For convenience we included samples of tables in data/. These are just samples, not the full tables. For the sake of example and so you can reason about the characteristics of the samples you got, we included example/crawl.py which is the script that generated the follows sample (and corresponding profiles). Feel free to generate your own sample by modifying example/crawl.py.

Deploying your bot

For development, it's simplest to just run the bot locally (python3 main.py). When you want to deploy it for real, there are many free or low cost cloud hosting options like Heroku or Fly.io.