ReAct, planner-executor, tool use, memory, multi-agent and MCP — build a Claude-Code/Codex-style coding agent end to end.
An agent is a loop — model, tools, memory, stop condition. Build the ReAct loop by hand in ~40 lines, watch it run live, then learn exactly when interviewers expect you to reach for it versus a plan-execute or a plain workflow.
When one ReAct loop isn't enough — plan once then execute, fan work out to an orchestrator and workers, and hand off between specialist agents. Plus the honest cost math on multi-agent systems and the senior heuristic for when they actually help.
Tools are the agent's hands — and the agent-computer interface (ACI) is as much a design surface as a human UI. Master function-calling mechanics, the bash-vs-dedicated-tool decision, and MCP, the protocol that turned tools into a plug-in ecosystem. Then assemble the tool surface of a Codex/Claude-Code-style coding agent.