Complex agent frameworks like LangGraph are becoming obsolete because models can now orchestrate themselves using a single system prompt.
The current trend in AI development is building intricate software layers to manage agent workflows. This research proves that self-orchestration inside a prompt actually beats these external frameworks in success rates. Models are now intelligent enough to handle their own routing and tool selection without outside help. This simplifies the development stack by removing the need for heavy orchestration libraries. It suggests that most of the complexity we are building around AI is already unnecessary. Developers should stop building cages for their agents and start writing better instructions.
In-Context Prompting Obsoletes Agent Orchestration for Procedural Tasks
arXiv · 2604.27891
Agent orchestration frameworks -- LangGraph, CrewAI, Google ADK, OpenAI Agents SDK, and others -- place an external orchestrator above the LLM, tracking state and injecting routing instructions at every turn. We present a controlled comparison showing that for procedural tasks, this architecture is dominated by a simpler alternative: putting the entire procedure in the system prompt and letting the model self-orchestrate. Across three domains -- travel booking (14 nodes), Zoom technical support