Agent patterns
Proven interaction structures you can implement in any stack. Each pattern includes when to use it, failure modes, and evaluation notes.
Human-in-the-Loop Pattern
intermediateHITL patterns pause agents for human approval, clarification, or takeover on risky or ambiguous steps.
Memory Systems Pattern
intermediatePatterns for agent memory: working state, summaries, long-term stores, and retrieval with governance.
Multi-Agent Pattern
advancedMulti-agent systems split work across specialized agents with defined roles, handoffs, and shared state.
Plan-and-Execute Pattern
intermediatePlan-and-execute agents draft a multi-step plan first, then carry out steps with optional re-planning.
ReAct Pattern
beginnerReAct interleaves reasoning traces with tool actions so agents can think, act, and observe in a loop.
Reflection and Self-Critique Pattern
intermediateReflection patterns add a critique step so agents can check work, catch errors, and improve outputs before finalizing.
Routing and Handoffs Pattern
intermediateRouting patterns send tasks to the right agent, model, or workflow based on intent, risk, and skills.
Tool Use Pattern
beginnerHow agents call tools reliably: schemas, selection, error handling, permissions, and verification of side effects.