## Principles to follow when enriching the logging context

### Don't add values coming from a scope outside of the current method

If a method receives a value as a parameter, it should trust that, if it was relevant, the caller already added this value to the context. Add values to the context as upstream as possible in the call chain. This ensures to not add duplicates, possibly using slightly different names to track the same value. The context was implemented to ensure that values did not have to be passed down the entire call stack to be logged at relevant points.

### Don't dump string representations of complex objects

The purpose of the context is to be consumed by structured logging frameworks. Dumping the string representation of an object, like a Scala case class instance, means embedding some form of string formatting in another (likely to be JSON). This can be difficult to manage and parse, so stick to simple values (strings, numbers, dates, etc.).

## Principles to follow when enriching the logging context

### Don't add values coming from a scope outside of the current method

If a method receives a value as a parameter, it should trust that, if it was relevant, the caller already added this value to the context. Add values to the context as upstream as possible in the call chain. This ensures to not add duplicates, possibly using slightly different names to track the same value. The context was implemented to ensure that values did not have to be passed down the entire call stack to be logged at relevant points.

### Don't dump string representations of complex objects

The purpose of the context is to be consumed by structured logging frameworks. Dumping the string representation of an object, like a Scala case class instance, means embedding some form of string formatting in another (likely to be JSON). This can be difficult to manage and parse, so stick to simple values (strings, numbers, dates, etc.).