Skip to main content
Act Operator is a CLI and project standard for building LangGraph 1.0+ applications with a clear, modular architecture. If you are coming from the intro, this page explains what the tool provides and how the structure fits together. It gives you:
  • A standardized project model called an Act
  • A monorepo-friendly layout for multiple graphs (Casts)
  • Built-in Skills for AI-assisted architecture, development, and testing
  • Explicit context via CLAUDE.md files so tools and teammates stay aligned

​
Why this exists

LangGraph makes it easy to build agentic workflows. What it does not give you is a standard way to organize them as the project grows. Act Operator makes that structure explicit so graphs remain understandable, testable, and evolvable.

​
How it works

  • act new scaffolds an Act project with a consistent layout
  • Each Cast is an independent graph package with its own state, nodes, and dependencies
  • Skills guide architecture → implementation → testing workflows
  • CLAUDE.md captures decisions so AI tools do not guess

​
Key concepts

Act The standardized project structure for a LangGraph application. Cast An independent graph package inside an Act project. Skill Instruction folders that AI tools can discover and follow. CLAUDE.md Design context and specifications for the project and each Cast.

​
Quick start

Requires Python 3.11+. 
uvx --from act-operator act new

​
Next steps

Install

Install and setup details

QuickStart

Go from zero to a working cast.

Template Architecture

Understand the generated structure.