agents like to read a lot to gather context before and after working, so optimizing for reading makes a big difference!
if you're building agents, i highly recommend reading this (or giving it to your agent and ask it to implement the findings)
Optimizing reading for AI agents cuts context-gathering costs
AISummary
Eric Zakariasson argues that agents spend heavily on reading context before and after work, so optimizing that reading makes a major difference. He recommends the linked guide to builders, or handing it to an agent to implement its findings.
Cursor's related post reports 7% lower token costs with no drop in agent quality, achieved through tighter prompts, selective tool loading, better caching, and compressed file reads.
Post on XView on X
@ericzakariasson
We've reduced token costs in Cursor by 7% with no drop in agent quality. Savings came from tighter prompts, selective tool loading, better caching, and compressed file reads.
Source: eric zakariasson · x.comPublished · added here