techsd / all-your-github-are-belong-to-us

:octocat: Save all your GitHub data to one place, private & public. A Webhook.

all your github are belong to us

GitHub is a fantastic tool — I use it constantly. I'd like to understand more of how I work on GitHub, but I don't have all my data. I can setup an RSS hook on my public feed, but I do a lot of work on private repos as well.

This webapp is a simple data collector: add it to repos in question and it'll log everything you do on that repo into a MongoDB database. Each collection is named after the event type. Pretty cool, huh?

Installation

Simple, just run script/bootstrap from your clone.

You might need to create a token. Run script/generator and follow the necessary prompts. Keep a record of this token — you'll need it for later configuration. Hell, save it to a .env file, why don't ya?

Deployment

I use Heroku because I'm obsessed with how easy it is to use. I also use hk because Go is freaking awesome. Download hk then run the following:

$ hk create my-github-log-app-thingy
$ hk addon-add mongohq
$ hk set GITHUB_USERNAME=your-username
$ hk set GITHUB_TOKEN=abc123
$ hk set AYGABTU_URL=http://my-github-log-app-thingy.herokuapp.com

Then you'll have to tell your aggregator which events you care about. You can view all the events here. Now, take the events you want and construct a JSON array. Then, set a string'd version of that as your GITHUB_HOOK_EVENTS:

$ hk set GITHUB_HOOK_EVENTS='["push","pull_request"]' # just an example!

Then deploy your code:

$ git push heroku master

Fetching data

It's all in Mongo. Now it's your job to get and understand it.

API endpoints are still in progress. Wanna help? Check out issue #2.