Does it matter how you feed an agent its context?

Our short summary of OLX’s engineering write-up. The full story, with all the detail, is on their blog.

Problem

Engineers ran out of token budget while using a coding agent for a complex backend task.

They compared four ways to give the agent context—single prompt, handoff notes, etc.—and measured cost and performance for the same coding job.

Want the architecture, code and numbers? They’re in the original post.