CoffeeBot

Your friendly neighborhood Coffee & Bagel meeting generator

This bot was created to post 1 on 1 (or more) random pairs for meetings where users can talk about life and things while enjoying some coffee or breakfast pastry. It tries to avoid the same match up over the past nCr meeting generations. Slack users included in these meetings are filtered by EMAIL_DOMAIN which can be configured - that way single channel guests or what have you are not included.

bagelbot is a Slack bot written in python that connects to slack via the RTM API. To generate a meeting and post it to Slack, you'll need a Slack API token. Add @bagelbot as a bot to your custom integrations at https://slack.com/apps/manage/ under Custom Integrations then Bots. You need to add a "classic" app, since Slack's API changed in the meanwhile.

This page can be used to create Slack application: https://medium.com/xandr-tech/bagel-bot-reimaging-donut-1fa68154a80c

Screenshots

Meeting Pairings posted to Slack Channel

Checking for Attendance

Dependencies

python-slackclient

Installation

There are a couple ways to run this project. First way, you can just check out the code and install a local virtualenv and run the scripts. The other is to use the provided Dockerfile to run commands.

Download the bagelbot code

git clone git@github.com:zooshgroup/bagelbot.git
cd bagelbot

Configure bagelbot by editing the config.py file accordingly (use your email domain, slack token, etc.)

Local Development

Install dependencies (virtualenv is recommended.)

pip install -r requirements.txt

Docker

docker build -t bagelbot .

Running

For syncing storage to S3, set AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY and AWS_DEFAULT_REGION environment variables.

Run locally

You can generate a meeting at any time, using generate_meeting.py.

./generate_meeting.py

Use the -h option for optional arguments. If you want to schedule this job in say a crontab somewhere, it would look like this:

Run check_attendance.py ahead of your meeting (the default time limit on the attendance check is 15 minutes). This script will run for the entirety of that time limit listed in config.py or as soon as all Slack users have responded.
After that time limit, say 15 minutes later, schedule generate_meeting.py to run. If there's an upcoming meeting in the shelf storage, and the --force-create option is passed, a meeting will be generated, sent out to the configured slack channel, and stored into the history key (a list of past meetings) of the shelf.

Run with Docker

You can run the individual scripts locally like above, or using a docker image such as:

docker run -e AWS_ACCESS_KEY_ID -e AWS_SECRET_ACCESS_KEY -e AWS_SECRET_ACCESS_KEY -e AWS_DEFAULT_REGION -it bagelbot python check_attendance.py --s3-sync --users ben

If you want to run Bagelbot as a Service (BaaS), you can use service.py to do so. This script checks to see if attendance should be checked at a certain time and the same with meeting generation. See config.py for an example of meeting times and frequencies. If S3_BUCKET is set, the SHELVE_FILE will be uploaded to S3 upon every operation that would change the state of the file.

docker run -e AWS_ACCESS_KEY_ID -e AWS_SECRET_ACCESS_KEY -e AWS_SECRET_ACCESS_KEY -e AWS_DEFAULT_REGION -it bagelbot

Development

There is a Makefile provided that uses pyenv-virtualenv to manage a python 3.6.5 virtual environment. If you have pyenv & pyenv-virtualenv installed properly (refer to their respective readme's), then you just need to run:

make install-dev

Run in production

Steps to run in "production:

create a t3.nano VM with read/write access to an S3 bucket
create an S3 bucket to store the storage file, upload a dummy storage file
place any secret settings in config_private.py
build the image, upload to ECR
install docker, check out the docker image from ECR and run it as a service

kohend/bagelbot