At Stripe, engineers now collaborate on crafting the perfect prompt to guide AI agents. This new form of teamwork focuses on articulating the problem clearly and providing the right context, rather than co-writing code line-by-line. This can involve other engineers, data sources, or even other agents.

A single LLM struggles with complex, multi-goal tasks. By breaking a task down and assigning specific roles (e.g., planner, interviewer, critic) to a "swarm" of agents, each can perform its bounded task more effectively, leading to a higher quality overall result.

To build a useful multi-agent AI system, model the agents after your existing human team. Create specialized agents for distinct roles like 'approvals,' 'document drafting,' or 'administration' to replicate and automate a proven workflow, rather than designing a monolithic, abstract AI.

Despite extensive prompt optimization, researchers found it couldn't fix the "synergy gap" in multi-agent teams. The real leverage lies in designing the communication architecture—determining which agent talks to which and in what sequence—to improve collaborative performance.

Before delegating a complex task, use a simple prompt to have a context-aware system generate a more detailed and effective prompt. This "prompt-for-a-prompt" workflow adds necessary detail and structure, significantly improving the agent's success rate and saving rework.

Instead of immediately asking an AI to perform a complex task, first prompt it to create a functional spec or a sequential plan. Go back and forth to align on this plan before instructing it to execute, which significantly improves the final output's quality and relevance.

An AI coding agent's performance is driven more by its "harness"—the system for prompting, tool access, and context management—than the underlying foundation model. This orchestration layer is where products create their unique value and where the most critical engineering work lies.

Separating AI agents into distinct roles (e.g., a technical expert and a customer-facing communicator) mirrors real-world team specializations. This allows for tailored configurations, like different 'temperature' settings for creativity versus accuracy, improving overall performance and preventing role confusion.

The most powerful AI systems consist of specialized agents with distinct roles (e.g., individual coaching, corporate strategy, knowledge base) that interact. This modular approach, exemplified by the Holmes, Mycroft, and 221B agents, creates a more robust and scalable solution than a single, all-knowing agent.