Distinguishing various steps

[ajram23]

Describe the feature or potential improvement

At the moment, it's really hard to distinguish various steps in a trace when everything is a span. Perhaps it's because I am using langchain. It would be really helpful to identify visually (either through icons etc) between tool calls, llm calls, prompts, code functions and output parsers like pyndatic. Also what does the badge and % icons do? (Tooltips would be helpful)

Additional information

No response