Running through docker?

Worked but then had issues? try docker-compose down --remove-orphans (This will cause all data to be lost)

Running locally?

No CRUD operations were requested hence I didn't use any serializers
For the client you can use any browser and swagger should be bundled as part of the site
Utilized Django Admin to veirify data through UI
I created a database migration that creates an admin user to login to admin site (user/pass: admin/admin123)
Every task status can be tracked by fetching the status of the repository

I did not adhear to th pep 8 line width for easier coding (and because I had issues with flake8).
No testing but that can be done uppon request including a mock server.
For simplicity I did not make credentials env vars but that is best practice!
Since there was no requirement for operations on the DB I didn't use the ModelViewSet.
For the same reason as above, I didn't use an API router to automatically handle the urls.
Because of time constaint reasons I did not handle malicious and non-valid api calls (including 4**, 5**)
Because of time constaint reasons I did not implement propper logging
In real life project I like to organize an "apps" folder and a config folder as I don't like the default layout
in real life since it's a many to many and not a dynamic field, cache should be used for all RepoRequirements and adding a flag reset on request (update=true), this can be done by using a custom model manager and caching the specific query
The API requests should be POST (not GET) but this way it's easier to launch through the browser
I would've liked to spend another hour or two for setting up an email that gets sent each time a signal has fired off for a status change of a RepoItem
All of the above would likely take me a long time but if it's the difference between pass and fail please let me know and I'll upgrade the project