Remove unnecesarry properties when parsing any results from a GitHub API request

Right now, when we make an HTTP request to the GitHub API, we are parsing all of the properties from the response, even the ones that we are not using.

Sometimes, some properties will have unexpected values (like `null` values), in which case the flow will fail.
We should only parse for the properties we actually need.

For this issue, we have to look at all the places we are making those calls and then change the parse step accordingly.

Start by locating every HTTP call to the GitHub API and its response-parsing step across the repository. For each call, identify which response properties the surrounding code actually uses and verify how unexpected null values currently fail. The work is done when unused properties are no longer parsed and all affected flows still handle the needed values.

Written by the indexing model from the issue text.