# git-bob

First, we start by creating a new Github issue. Navigate to our git-bob-test repository where the world happiness data is stored for you and [create a new issue](), e.g.: 

```
You find 2015.csv, 2016.csv, ... 2019.csv and there are columns such as "Country", "Happiness Rank". 

I would like to plot the happiness rank of neighboring countries of Germany and Germany over time.

git-bob try to solve this
```

**Note:** 
* It is required to specify the format "I need the code as Jupyter Notebook."
* The last line starting with `git-bob try to ...` is required to trigger the AI agent to do something.


![alt text](create_issue.png)

After a moment, you will be notified that git-bob is approaching the problem:




Finally, it will respond. The response will look like it was written by human (the human who configured git-bob):

![alt text](git-bob-response.png)

If the result is not as expected, you can ask git-bob to make modifications. And keep in mind, you need to ask `git-bob try to  ...` to trigger it.

If you're happy with the propose code and visualization, you can ask the AI to  send a pull-request. In this case you  simply instruct `git-bob implement ...`. [Read  more about  git-bob`s trigger words](https://github.com/haesleinhuepf/git-bob#usage-trigger-words).


![alt text](git-bob-implemented.png)

And this is how the final pull-request may look like:

![alt text](git-bob-pr.png)