Alibaba just dropped Qoder 1.0, and it's no longer just an AI-powered code editor. The company is repositioning it as an "intelligent agent development workbench," with the core Quest mode now running as a standalone window alongside the traditional code editor. Developers set a goal, and the agent takes over — executing, validating, and delivering the entire task flow.
For multitaskers, Qoder 1.0 lets you run agent tasks across multiple code repositories simultaneously from a unified panel. When a task wraps up, it automatically generates a delivery manifest with documentation and code changes for human review.
But the headliner is the new team-level knowledge engine. It consolidates what used to be scattered developer habits, project architecture, and tech stack specs into a shared repository. Alibaba's own data shows code retention rates jumped 11% and conversation rounds dropped 33% after the engine went live.
Also new: an expert panel — planning, coding, testing — is now baked into the Quest window. Developers can assemble custom agent teams, equipping each with domain knowledge and external tool integrations.