Debian based Customizable Hubot Container – 779 MB – Updated 01/09/2016

This container is built from node:4.2.4
Any of the examples below can pull off of the Quay registry by putting in the front of the namespace/repository
This build must run off of nothing higher than node 4.2, running the build on node 5 / npm3 breaks dependancies (stringprep, hubot-hipchat)


Hubot is your company’s robot. Install him in your company to dramatically improve and reduce employee efficiency.

GitHub, Inc., wrote the first version of Hubot to automate our company chat room. Hubot knew how to deploy the site, automate a lot of tasks, and be a source of fun in the company. Eventually he grew to become a formidable force in GitHub. But he led a private, messy life. So we rewrote him.

Today’s version of Hubot is open source, written in CoffeeScript on Node.js, and easily deployed on platforms like Heroku. More importantly, Hubot is a standardized way to share scripts between everyone’s robots.

Official Hubot Site:


This container is built from official node:4.2.4. The image resides on both the docker hub as well as the quay registry. If you would prefer to use the images distributed from, just simply append to the front of the namespace/repository designation as such in the examples

This containers purpose is to get a customizable hubot instance up and running with a single docker run statement. The container is built with environment variables that allow the user to plug in custom values which will allow the container to configure itself on first run. This will give the user a fully customized experience just as if you set up hubot on your own from scratch. The image has the gtalk, slack and hipchat adapters installed to allow integration into these 3 main chat systems.


You Tube Demo:

Coming Soon!


Container Variables:

The container is built to allow several configurable variables to be passed in at run time. The values are as follows:

Standard Options

  • OWNER – This is the user on the system that owns the hubot binary / directories. It should run off the hubot user.
  • HUBOT_NAME – This is the name that you choose for your hubot chat robot
  • DESCRIPTION – Quick description of what your robot is or what it will do.
  • ADAPTER_NAME – This is the adapater that hubot will use to connect to your chat service. gtalk, slack, and hipchat are built in
  • WHITELIST_DOMAINS – Domain list allowed to connect to the listening hubot interface on the hubot server.
  • TERMTAG HUBOT – Environment variable to hold the highlighted bash shell tag. Currently set to the repo name, but can be change from within the container via TERMTAG=”Whatever You Want”

Slack Options

  • HUBOT_SLACK_TOKEN – The token generated on slack that hubot will use to connect to your slack service
  • HUBOT_SLACK_TEAM – The team that hubot will be part of when connected to your slack service. This is your team name.
  • HUBOT_SLACK_NAME – The name of the slack user that hubot will bind to in your chat service.

Google Options

  • GTALK_API_KEY – The token generated on google that hubot will use to connect to your gtalk service

Hipchat Options

  • HUBOT_HIPCHAT_TOKEN – The token generated on hipchat that hubot will use to connect to the hipchat service
  • HUBOT_HIPCHAT_JABBERID – The user id ( identifier that tells hubot what hipchat user to bind to
  • HUBOT_HIPCHAT_NAME – The name of the hipchat user that hubot will bind to in your chat service.
  • HUBOT_HIPCHAT_PASSWORD – The password of the created hipchat user that hubot will use to connect to the hipchat service
  • HUBOT_HIPCHAT_ROOMS – The room listing that hubot will be added to in your hipchat service.


Running the Container to connect to slack:

Log into your slack service, and navigate to ”’Configure Apps”’. Search for and Install the Hubot Integration. Once installed click on the hubot app, and click the edit button. This screen will easily allow you to set all of the parameters required for your hubot integration.


docker run -d -it \
--name hubot -h hubot \
--restart always \
-p 8080:8008 \
-e ADAPTER_NAME=slack \
-e HUBOT_SLACK_TOKEN=paste_token_here \
-e HUBOT_SLACK_TEAM=paste_your_team_here \
-e HUBOT_SLACK_NAME=paste_slack_hubot_user_here \


Running the Container to connect to hipchat:

Log into your hipchat service, and navigate to ”’Account Settings”’. On the Account Settings Page Go to the bottom of the left menu pane, and click on ”’XMPP/Jabber Info”’ This will show your hipcat Jabber ID in the form of Copy the first 5 digits, as this is your team hipchat ID.


Next, Click the people tag and locate your hubot user. Once you find the user click on them, and examine the URL, it will show
This number at the end of the URL string is the hipchat user ID. this will be combined with your team ID to form the Hubot Jabber ID (JID) that hubot will use to connect to the hiphcat service.


Next Click navigate to as the admin user, this will show all of the authorized API Auth Tokens that your comapany allows to connect to the hipchat service. (ALL assuming that this has been set up already). Here is where you can find the Hipchat Auth Token for hubot.


docker run -d -it \
--name hubot -h hubot \
--restart always \
-p 8080:8008 \
-e ADAPTER_NAME=hipchat \
-e HUBOT_HIPCHAT_TOKEN=paste_token_here \
-e HUBOT_HIPCHAT_JABBERID=paste_your_hipchatuser_id@chat.hipchat.com_id_here \
-e HUBOT_HIPCHAT_NAME=paste_slack_hubot_user_here \
-e HUBOT_HIPCHAT_PASSWORD=paset_hipchat_user_password_here \

These examples will assume the following

  • OWNER = hubot
  • HUBOT_NAME = hubot
  • DESCRIPTION = “Company Chat Robot”

Setting any of the above values will tell the container to replace the default values already set within the container with the values that are supplied at runtime. For example, if you pass in -e HUBOT_NAME=tester then hubot will be pre named and respond as tester instead of hubot.

These examples will start a new container named hubot. The restart policy will be set to always, meaning that if the container crashes unexpectedly, it will automatically kick itself back off. At this point your app is fully configured and you need just check your chat service to ensure that the hubot user has connected succesfully. Once connected you can test that the chat robot is working correctly by typing @hubot ping in a chat channel. Hubot should respond with the message Pong.


Included Scripts:

The following scripts are built into hubot with this build

  • hubot-scripts – Gives hubot the ability to respond to the ping/pong test as well as other default hubot actions such as help.
  • hubot-redis-brain – Allows hubot to communicate with a redis instance. If redis is not set up, it will run a local instance.
  • hubot-pugme – Extends hubot to send a pug image if you type @hubot pug me
  • hubot-rules – Extends hubot to send back robot rules by typing @hubot the rules
  • hubot-shipit – Extends hubot to send back a ship it squirrel meme by typing @hubot ship it
  • hubot-suggest – If you send hubot a PM with something it doesn not understand, it will offer suggestions.


Adding additional functionality scripts to hubot:

Attach to hubot by using docker attach hubot and hitting enter twice. Hit CTL C to kill the running hubot process. Navigate to the hubot folder /opt/hubot. Find the module for the script that you would like to add on the NPM site, and install the module by typing npm install hubot-script_name. Once npm installs the script, then edit the “external-scripts.json” file, and adding the newly installed script to the end of the list. Don’t forget to add a comma to the element above the new entry. Every element in the json array should have a value except for the last value.


Once completed, you can restart hubot by running bin/hubot --adapter $ADAPTER_NAME. That should run the hubot process again with the newly installed extention script. At this point you can exit the container. DO not type exit (which would shut down the container), but instead use the detach key combination of CTL P + CTL Q.


Launching the Container via docker-compose:

Copy the text below and paste it into a file named docker-compose.yml. Then you can navigate to the directory and provided that you have docker-compose installed, just issue the following command in order to launch off the application stack:

Remember to substitute your chat service values with the value examples in the compose syntax below before running it.

docker-compose up -d

NOTE: If you need assistance setting up docker-compose please visit to watch the tutorial or read about the installation process.


  image: appcontainers/hubot
  hostname: hubot
  stdin_open: true
  tty: true
  restart: always
  - "8080:8080"
  - OWNER=hubot
  - DESCRIPTION=Resistance is futile....
  - ADAPTER_NAME=slack
  - HUBOT_SLACK_TOKEN=1234-12345678910-abcd1234efghijklmn56opqr
  - HUBOT_SLACK_TEAM=your-team


  image: appcontainers/hubot
  hostname: hubot
  stdin_open: true
  tty: true
  restart: always
  - "8080:8080"
  - OWNER=hubot
  - DESCRIPTION=Resistance is futile....
  - ADAPTER_NAME=hipchat
  - HUBOT_HIPCHAT_TOKEN=1234567890123abcdefghijklmnopq
  command: /bin/bash

NOTE: The hipchat JID is a combination of your company identifier along with the hubot user id which can be found on the hipchat web site when you are logged in.


Dockerfile Change-log:

01/09/2016 - Hubot Container Created



Test Standalone Version
docker run -it \
--name hubot \
-h hubot \
-p 8080:8080 \
# Attach to container and navigate to /opt/hubot/bin
# This should give you the hubot prompt.