Small update to 'Introduction to Agents' docs format (#548)

docs/source/en/conceptual_guides/intro_agents.md

@@ -83,8 +83,8 @@ Until recently, computer programs were restricted to pre-determined workflows, t
 For some low-level agentic use cases, like chains or routers, you can write all the code yourself. You'll be much better that way, since it will let you control and understand your system better.
 
 But once you start going for more complicated behaviours like letting an LLM call a function (that's "tool calling") or letting an LLM run a while loop ("multi-step agent"), some abstractions become necessary:
-- for tool calling, you need to parse the agent's output, so this output needs a predefined format like "Thought: I should call tool 'get_weather'. Action: get_weather(Paris).", that you parse with a predefined function, and system prompt given to the LLM should notify it about this format.
-- for a multi-step agent where the LLM output determines the loop, you need to give a different prompt to the LLM based on what happened in the last loop iteration: so you need some kind of memory.
+- For tool calling, you need to parse the agent's output, so this output needs a predefined format like "Thought: I should call tool 'get_weather'. Action: get_weather(Paris).", that you parse with a predefined function, and system prompt given to the LLM should notify it about this format.
+- For a multi-step agent where the LLM output determines the loop, you need to give a different prompt to the LLM based on what happened in the last loop iteration: so you need some kind of memory.
 
 See? With these two examples, we already found the need for a few items to help us: