Fundamentals Rules

Learn why agent projects fail due to unbounded behavior. Discover how fundamental rules on scope, stopping, and context prevent cost spikes and unpredictable processes.

Condenses core agent fundamentals rules for scope, stopping, context, and model choice into a scannable quick-reference list.

Determine if an agent is the right solution for your problem. Learn when to use an agent versus a script, workflow, or single LLM call.

Learn rules for structuring, trimming, and refreshing agent context to ensure predictable behavior and prevent "model got dumber" issues.

Avoid common pitfalls in agent development. Learn to identify and fix recurring mistakes in building multi-turn, tool-using AI agents.

Learn ten best practices for AI agent design, covering scope, stopping conditions, context management, and model policy for robust agents.