Research shows that, similar to humans, LLMs respond to positive reinforcement. Including encouraging phrases like "take a deep breath" or "go get 'em, Slugger" in prompts is a deliberate technique called "emotion prompting" that can measurably improve the quality and performance of the AI's output.
Implement human-in-the-loop checkpoints using a simple, fast LLM as a 'generative filter.' This agent's sole job is to interpret natural language feedback from a human reviewer (e.g., in Slack) and translate it into a structured command ('ship it' or 'revise') to trigger the correct automated pathway.
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 agent development process can be significantly sped up by running multiple tasks concurrently. While one agent is engineering a prompt, other processes can be simultaneously scraping websites for a RAG database and conducting deep research on separate platforms. This parallel workflow is key to building complex systems quickly.
Instead of manually refining a complex prompt, create a process where an AI agent evaluates its own output. By providing a framework for self-critique, including quantitative scores and qualitative reasoning, the AI can iteratively enhance its own system instructions and achieve a much stronger result.
To explain the LLM 'temperature' parameter, imagine a claw machine. A low temperature (zero) is a sharp, icy peak where the claw deterministically grabs the top token. A high temperature melts the peak, allowing the claw to grab more creative, varied tokens from a wider, flatter area.
When building multi-agent systems, tailor the output format to the recipient. While Markdown is best for human readability, agents communicating with each other should use JSON. LLMs can parse structured JSON data more reliably and efficiently, reducing errors in complex, automated workflows.
To build truly effective agents, adopt a "founder's level of service" mindset. This involves an intensive discovery process to become a temporary expert in the client's business, culture, and brand voice. This deep, meticulous care ensures the final AI system is perfectly aligned with the client's intentions.
Anthropic's Claude models are specifically trained on XML. By structuring system instructions using XML tags (e.g., <role>, <instructions>), you align with the model's training data. This provides better organization and can unlock additional functionality and more reliable outputs compared to using plain text prompts.