# Plugins and Continuous Integration Testing

**URL:** https://meta.discourse.org/t/plugins-and-continuous-integration-testing/49757
**Category:** Development
**Created:** [September 6, 2016, 12:44am UTC](https://meta.discourse.org/t/plugins-and-continuous-integration-testing/49757 "2016-09-06T00:44:57Z")
**Posts on this page:** 1
**Page:** 1

<div class="post-metadata">

### Author: ![rimian](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/rimian/32/120658_2.png) [@rimian](https://meta.discourse.org/u/rimian)
#### Post date: [September 6, 2016, 12:44am UTC](https://meta.discourse.org/t/plugins-and-continuous-integration-testing/49757/1 "2016-09-06T00:44:57Z")

</div>

I’m looking at extending an existing plugin but it has no tests. I’d rather code to a TDD cycle but I cannot find much documentation on how to build plugins this way.

Here’s an example of a plugin using `Grunt` as the task runner:  
[https://github.com/chrisbeach/discourse-plugin-stripe](https://github.com/chrisbeach/discourse-plugin-stripe)

Of course core uses a rake task which runs `phantomjs`:

```
cmd = "phantomjs #{test_path}/run-qunit.js http://localhost:#{port}/qunit"

```

There’s also `Gulp` and I am thinking about installing discourse and running my plugins inside of that (as above) but that would be slow…

Is there a convention for running QUnit specs for a plugin on Travis?