@@ -23,9 +23,9 @@
 
 ## Single-turn prompts
 
-Instead of messaging the agent in an ever-growing conversation, you first select task-relevant files, then send [lean markdown prompts](#prompts) with chatbots, API calls or headless CLI invocations.
+Instead of messaging the agent in an ever-growing conversation, you first select task-relevant files, then send [lean markdown prompts](#prompts) with chatbots, API calls, or headless CLI invocations.
 
-Because they make intent very easy to unpack, models _think_ less lowering response times and saving quotas 10x!
+**LLMs focused on the task from the start _think_ less, lowering response times and saving quotas 10x!**
 
 > [!TIP]
 >
Original file line number	Diff line number	Diff line change
`@@ -23,9 +23,9 @@`
`23`	`23`
`24`	`24`	`## Single-turn prompts`
`25`	`25`
`26`		`-Instead of messaging the agent in an ever-growing conversation, you first select task-relevant files, then send [lean markdown prompts](#prompts) with chatbots, API calls or headless CLI invocations.`
	`26`	`+Instead of messaging the agent in an ever-growing conversation, you first select task-relevant files, then send [lean markdown prompts](#prompts) with chatbots, API calls, or headless CLI invocations.`
`27`	`27`
`28`		`-Because they make intent very easy to unpack, models _think_ less lowering response times and saving quotas 10x!`
	`28`	`+LLMs focused on the task from the start _think_ less, lowering response times and saving quotas 10x!`
`29`	`29`
`30`	`30`	`> [!TIP]`
`31`	`31`	`>`