Receive Pull Requests notifications from Bitbucket and send them to Slack.
As for 2014-09-26 we support all BitBucket's action types:
- created
- updated
- approve
- unapprove
- declined
- merged
- comment_created
- comment_deleted
- comment_updated
See Bitbucket's official docs: Pull Request POST hook management
Bitbucket does not provide all necessary data in actions responses. For example, for a comment_*
action we don't have the PR number, link nor name.
So all we can do is something like "Comment posted for a PR" and then the snippet of the comment. But currently there's no way to know where did this comment came from.
- Bitbucket repository with admin rights
- Slack Incoming Webhook Url: Get your Slack token from your "integrations" page
- Node.js OR Docker
The configuration variables are set with environment variables and/or using .env
file (environment variable takes preference over .env
file if found).
Environment Variable | Required | Description | Example |
---|---|---|---|
SLACK_WEBHOOK | Y | The incoming hook url found on your Slack team's integration page | https://hooks.slack.com/services/XX/XXX/XXXX |
SLACK_USERNAME | N | Username of the Slack bot. If not set, bot will default to integration settings. | BitbucketNotification |
SLACK_CHANNEL | N | Channel to post notifications on. If not set, bot will default to integration settings. | RepositoryUpdate |
HEX_INFO | N | Hex color of updated, and created | #3498db |
HEX_DANGER | N | Hex color of declined | #e74c3c |
HEX_WARNING | N | Hex color of unapprove, comment: created, comment: deleted, and comment: updated | #f1c40f |
HEX_SUCCESS | N | Hex color of merge, and approve | #2ecc71 |
If you want to use .env
file, copy the example.env
as .env
and modify it as needed.
Your configuration would look like the example below:
PORT=5000
SLACK_WEBHOOK=https://hooks.slack.com/services/A123ka9/A123910a9d8/mkas929199sad83lmk7h
SLACK_USERNAME=AwesomeBot
SLACK_CHANNEL=Repository
HEX_SUCCESS=#2ecc71
HEX_DANGER=#e74c3c
Note: Setting the SLACK_USERNAME
or SLACK_CHANNEL
will override the settings set on the incoming webhook integration page. If you want your team to edit any of these settings without redeploying, do not add them to your .env
file.
Important: if you're going to use a .env
file AND using Docker, edit it before building the Dockerfile.
When running the service in Docker container, the config values can be provided as parameters:
# Starts Docker container in daemonized mode
docker run -e PORT=5000 -e SLACK_TOKEN=123123 \
-e SLACK_DOMAIN=company -e SLACK_CHANNEL=channel \
-p 5000:5000 -d bitbucket-slack-pr-hook
- Set up a server address in your local infrastructure that will serve this application (eg:
slackapi.mycompany.com
orslackapi.heroku.com
) - Clone/download this repo to your chosen server
- Configure your application according to the "Configuration" section above
- Install NodeJS if you don't have it
- Run
npm install
in the app's root folder
Important note: make sure you don't have any firewall blocking the incoming TCP port (default is PORT 5000 as defined in the "Configuration" section above)
Service can also be installed & deployed using Docker containers, which makes it easy to setup the environment without worrying about the requirements.
-
Clone/download this repo
-
Install Docker (using system packages in Linux, in Windows or Mac OS X you can use Boot2docker)
-
Start Docker service (or
boot2docker up
) -
Build the Docker image:
docker build -t bitbucket-slack-pr-hook .
-
Start container with appropriate
-e
config parameters:docker run -e PORT=5000 -e SLACK_WEBHOOK=webhookurl
-p 5000:5000 -d bitbucket-slack-pr-hook -
Ensure the container is running (you should also be able to access the service using web browser:
http://<dockerhost>:5000/
).Note: In Linux the
<dockerhost>
islocalhost
, with Boot2docker use the IP reported by the command:boot2docker ip
- In your main Bitbucket repository, go to Settings > Hooks and create a new
Pull Request POST
hook - Set up the URL as
http://<server>:<port>{/<channel>}
. *<server>
is your host FQDN or its IP address *<port>
is either 5000 or any other you defined in the configuration section *<channel>
is an optional Slack channel where you want to receive this specific notifications - if it's not defined here it will use the one you defined in Configuration -section.
- Run
npm start
(ornode server.js
) to fire up the application (you can donode server.js &
to run it as a daemon in your Linux box)
- Run
npm run build-container
to build the container - Run
npm run start-container
to start the container and the server inside it - When needed, you can use
npm run stop-container
andnpm run reload-container
This repo was inspired by: https://github.com/kfr2/bitbucket-pull-request-connector