Layer 2: Creation & Orchestration — The Agentic Economy

Layer 2 of 7 — The Agentic Economy
Creation & Orchestration
Build, orchestrate, and monitor agentic workflows
L1Agents
L2Creation
L3Platforms
L4Models
L5Knowledge
L6Compute
L7Physical
The tools and frameworks for building AI agents, orchestrating multi-agent systems, and evaluating their performance.
29
companies
3
subcategories
Creator
→ Runtime → Eval
Companies in this layer
Creator Tools
Bolt.newClaude CodeCursorFigmaGitHub CopilotLovableReplitRobloxSourcegraphv0Windsurf
Orchestration / Agent Runtime
A2AAgent FrameworkCloudflare AgentsCrewAIGoogle ADKLangChainLangGraphLlamaIndexMCPn8nOpenAI Agents SDKTemporalVertex AI Agent EngineZapier
Observability & Eval
Arize AIBraintrustLangfuseLangSmith

Layer 2: Creation & Orchestration is the builder layer of the Agentic Economy — the tools, frameworks, and runtimes that developers use to create, coordinate, and monitor AI agents.

Creator Tools

AI-powered development environments are transforming how software gets built. Cursor, Windsurf, and GitHub Copilot provide AI-assisted coding. Claude Code enables fully agentic software engineering. Replit, Bolt.new, and Lovable let non-developers build complete applications through natural language. Figma brings AI to design, while Roblox uses AI for 3D world creation.

Orchestration & Agent Runtime

Multi-agent systems need coordination layers. LangChain, LangGraph, and LlamaIndex provide the most widely adopted open-source frameworks. CrewAI specializes in multi-agent collaboration. MCP (Model Context Protocol) and A2A (Agent-to-Agent Protocol) define how agents connect to tools and communicate with each other. n8n, Zapier, and Temporal handle workflow automation and durable execution.

Observability & Evaluation

As agents become autonomous, understanding their behavior is critical. Arize AI, Langfuse, LangSmith, and Braintrust provide the monitoring, tracing, and evaluation tools that let teams understand what their agents are doing and how well they're performing.