10 — AI Agents¶
AI agents: architecture, tools, planning, memory, routing, multi-agent systems, and evaluation.
Scope¶
Cheat sheets for building agentic systems. Single-turn LLM usage belongs to 08-LLM, retrieval to 09-RAG.
Cheat Sheets¶
| Sheet | Covers |
|---|---|
| ai-agents.md | Agent loop, architecture, patterns, failure modes, human-in-the-loop |
| tools.md | Tool design, schemas, categories, validation, guardrails |
| tool-calling.md | OpenAI/Anthropic/Gemini tool calling, execution loops, error handling |
| memory.md | Short-term, long-term, episodic, semantic memory, persistence |
| planning.md | ReAct, plan-and-execute, tree-of-thought, reflection |
| routing.md | Intent classification, conditional routing, handoffs, human gates |
| multi-agent-systems.md | Supervisor, workers, handoffs, debate, communication patterns |
| agent-evaluation.md | Metrics, failure modes, benchmarking, A/B testing |
Contribute¶
- Template: CHEATSHEET_TEMPLATE.md
- Guide: CONTRIBUTING.md
- Planned: CrewAI, AutoGen, LangGraph advanced patterns