The official Redux recommendation is to avoid testing thunks in isolation and instead opt for integration testing with redux connected components. The main reason given is that thunks are considered implementation details.

Why are thunks considered implementation details?


 In the context of async thunks in RTK can you give examples of how testing an async thunk might lead to false negatives and false positives in our tests? 



What are your opinions on testing thunks in isolation that conditionally dispatch other actions, something that my team at work often does.

• Example of a custom render method (docs on this)