Skip to content
Capture all links shared on slack public channels
Ruby HTML CSS JavaScript
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
app
bin
config
db
lib
log
public
spec
tmp
vendor/assets
.bowerrc
.env.example
.gitignore
.rspec
.ruby-version
Gemfile
Gemfile.lock
Guardfile
README.md
Rakefile
bower.json
config.ru

README.md

CircleCI

README

Remembrall is a vault for storing links. It was built for Slack, so it will capture and save all urls shared in the channels where you configured the hook.

It is composed by two parts:

  • API: the messages are posted by slack to a /message route. It can also be used by third-part systems as long as they respect the contract of the json (basically, the links in the text must be wrapped around <>)

  • Dashboard: by accessing the / route, you can visualize the saved urls and search the database.

Once Remebrall is up and running, all links shared in the chosen channels will automatically be saved. To know how to configure channels read the section Configuring the hook below.

A tip: duplicated links are never saved again!

Important info

  • Ruby Version: 2.3.0
  • Database: postgres (tested with version 9.4.5)
  • To run the test suite: bin/rspec

In development, copy the .env.example to a file named .env and add your slack token in it. You can generate a token here.

The database is configured in the config/database.yml file, and can also be modified with environment variables. Add any of these keys to your .env if you do not want to run with the default database settings:

  • POSTGRESQL_USER
  • POSTGRESQL_DATABASE
  • POSTGRESQL_PASSWORD
  • POSTGRESQL_HOST
  • POSTGRESQL_POOL
  • POSTGRESQL_PORT

For testing purposes, there is a seed to populate the first articles in the database. To populate it, just go to the console and run:

  bin/rake db:seed

Of course, it expects you to have already run bin/rake db:create db:migrate :)

Users

To fetch Slack users and save their ids and avatars you need to go to the console and run:

bin/rake import_slack_users

It will use the SLACK_TOKEN on the .env file and retrieve the information. You can run this command as many times as you need, the data is never destroyed. This is good because you can run it from times to times to ensure the avatars are in sync with slack.

Configuring the hook

Time to make it work. All you need to do is to go to your slack integrations page (https://.slack.com/apps/manage/custom-integrations), choose Outgoing Webhooks and Add Configuration.

Give it a name, choose a channel and fill the URL(s) field with this project url (it must be opened to the internet, localhost will not work because slack needs to communicate with that url) pointing to /message. Now, every time someone posts a message in the chosen channel Slack will post a json to us, that will be parsed in order to extract the urls.

You can repeat this configuration on as many channels as you want - but be aware that each channel will require a new Outgoing Webhook configuration.

Configuring the login with slack

The informaion about the integration is in the following url: https://api.slack.com/docs/oauth.

You can check your apps in https://api.slack.com/apps

The black list

There might be a list of urls you don't want to save. For example, there is no need to store hangouts' conversations or links to Remembrall itself. To prevent a link to be saved, include part of it in the BLACK_LIST constant inside the models/url.rb.

The name

The name is a homage to the Harry Potter Remembrall, a magical artifact created to remind its bearer that he/she has forgotten something. The idea behind this is that we kept forgetting the links we shared on slack, so now we can store them automatically, without having to take any special action. Once the link is shared, anyone can find it back easily.

Ngrok

In order to test remembrall on your local machine, you can use ngrok.

To run ngrok:

/Applications/ngrok http 80

Copy the forwarding url it will print on the terminal and use it on your webhook.

You can’t perform that action at this time.