Early agent development used simple frameworks ("scaffolds") to structure model interactions. As LLMs grew more capable, the industry moved to "harnesses"—more opinionated, "batteries-included" systems that provide default tools (like planning and file systems) and handle complex tasks like context compaction automatically.
Related Insights
A major trend in AI development is the shift away from optimizing for individual model releases. Instead, developers can integrate higher-level, pre-packaged agents like Codex. This allows teams to build on a stable agentic layer without needing to constantly adapt to underlying model changes, API updates, and sandboxing requirements.
Simply offering the latest model is no longer a competitive advantage. True value is created in the system built around the model—the system prompts, tools, and overall scaffolding. This 'harness' is what optimizes a model's performance for specific tasks and delivers a superior user experience.
The true building block of an AI feature is the "agent"—a combination of the model, system prompts, tool descriptions, and feedback loops. Swapping an LLM is not a simple drop-in replacement; it breaks the agent's behavior and requires re-engineering the entire system around it.
Early on, Google's Jules team built complex scaffolding with numerous sub-agents to compensate for model weaknesses. As models like Gemini improved, they found that simpler architectures performed better and were easier to maintain. The complex scaffolding was a temporary crutch, not a sustainable long-term solution.
The success of tools like Anthropic's Claude Code demonstrates that well-designed harnesses are what transform a powerful AI model from a simple chatbot into a genuinely useful digital assistant. The scaffolding provides the necessary context and structure for the model to perform complex tasks effectively.
The early focus on crafting the perfect prompt is obsolete. Sophisticated AI interaction is now about 'context engineering': architecting the entire environment by providing models with the right tools, data, and retrieval mechanisms to guide their reasoning process effectively.
"Context Engineering" is the critical practice of managing information fed to an LLM, especially in multi-step agents. This includes techniques like context compaction, using sub-agents, and managing memory. Harrison Chase considers this discipline more crucial than prompt engineering for building sophisticated agents.
The developer abstraction layer is moving up from the model API to the agent. A generic interface for switching models is insufficient because it creates a 'lowest common denominator' product. Real power comes from tightly binding a specific model to an agentic loop with compute and file system access.
An open-source harness with just basic tools like web search and a code interpreter enabled models to score higher on the GDPVal benchmark than when using their own integrated chatbot interfaces. This implies that for highly capable models, a less restrictive framework allows for better performance.
Salesforce's Chief AI Scientist explains that a true enterprise agent comprises four key parts: Memory (RAG), a Brain (reasoning engine), Actuators (API calls), and an Interface. A simple LLM is insufficient for enterprise tasks; the surrounding infrastructure provides the real functionality.
