Why Prompt Engineering might not be enough

TLDR; you might just want to go through the video at youtube The 7 Skills You Need to Build AI Agents

Above video discusses the seven skills required for proficiency in developing agents. These skills are:

System Design
Tool & contract design
Retrieval engineering
Reliability engineering
Security and safety
Evaluation and observability
Product thinking

Decoding the skills

System design is important because it helps you to design the overall architecture of your agentic system.
Tool & contract design is important because it helps you to design the tools and contracts that your agents will use to interact with the environment and other agents.
Retrieval engineering is important because it helps you to design the retrieval system that your agents will use to retrieve information from the environment and other agents.
Reliability engineering is important because it helps you to design the reliability system that your agents will use to ensure that they are reliable and can handle failures gracefully.
Security and safety is important because it helps you to design the security and safety system that your agents will use to ensure that they are secure and safe to use.
Evaluation and observability is important because it helps you to design the evaluation and observability system that your agents will use to evaluate their performance and observe their behavior.
Product thinking is important because it helps you to design the product that your agents will be used for and ensure that it meets the needs of the users.

If you are looking to build agentic systems, you should focus on developing these skills. Yes they cover a very wide spectrum of skills and yes you do not have to be an expert in all of these but having a decent understanding of them will help you to build better agentic systems. For what I have heard, evalation and observability is one of the most important skills to have in the current landscape of agentic development.