OpenAI and Cursor differ on who runs coding-agent coordinators
The New Stack says OpenAI's Agents API and Cursor Projects both split coding work between a coordinator and specialized agents, but differ on who runs the coordinators.
TLDR
The New Stack compares how OpenAI's Agents API and Cursor Projects use coordinators and specialized agents for coding work. It says the two approaches differ on who runs the coordinators, and examines what that means for context, permissions and reliability.
Combined views
535
1 Source, first seen 7h ago