LangGraph State Machines
Building stateful agent graphs:
- Define immutable graph state schemas.
- Route between agents using conditional edges based on tool outputs.
- Persist state checkpoints to database for resume-on-failure capabilities.
Design stateful multi-agent graphs, cyclical workflow execution, human-in-the-loop approvals, and checkpointing.
Andromity loads skills from folders you trust. Install with andromity skill install langchain-agent-workflows — autonomy and approval limits stay configurable per session.
# Install the "LangGraph Multi-Agent Workflows" skill You are installing an AgenticMarket skill into this workspace. Follow these steps exactly. ## 1. Detect your environment and choose the destination file - Andromity: `.andromity/skills/langchain-agent-workflows/SKILL.md` (or user scope `~/.andromity/skills/langchain-agent-workflows/SKILL.md`) - Antigravity / Gemini CLI: `.agent/skills/langchain-agent-workflows/SKILL.md` - Cursor: `.cursor/rules/langchain-agent-workflows.mdc` - Claude Code: append to `CLAUDE.md` (or `.claude/skills/langchain-agent-workflows.md`) - Windsurf: append to `.windsurfrules` - Cline / Roo Code: `.clinerules/langchain-agent-workflows.md` - GitHub Copilot: append to `.github/copilot-instructions.md` - ChatGPT / web only: paste the content into Custom Instructions - Any other agent: save to `.agent/skills/langchain-agent-workflows/SKILL.md` ## 2. Create the destination file - Preserve every existing file and configuration entry. - Create parent folders as needed. - Never overwrite unrelated content. ## 3. Write exactly this content ```markdown --- name: langchain-agent-workflows description: "Design stateful multi-agent graphs, cyclical workflow execution, human-in-the-loop approvals, and checkpointing." author: community license: mit version: 1.0.0 --- # LangGraph State Machines Building stateful agent graphs: - Define immutable graph state schemas. - Route between agents using conditional edges based on tool outputs. - Persist state checkpoints to database for resume-on-failure capabilities. ``` ## 4. Verify - Confirm the file exists and is readable. - Report the destination path and confirm the "LangGraph Multi-Agent Workflows" skill is installed and active. Skill metadata: - Slug: langchain-agent-workflows - Category: ai-agents - License: mit - Source: AgenticMarket
Building stateful agent graphs:
Design stateful multi-agent graphs, cyclical workflow execution, human-in-the-loop approvals, and checkpointing.
LangGraph Multi-Agent Workflows works seamlessly with Cursor (.cursor/rules), Claude Code (CLAUDE.md & .claude/skills), Windsurf (.windsurfrules), Google Antigravity (.agent/skills), Andromity (.andromity/skills), and Cline.
Click '1-Click Prompt Install' in the right sidebar or copy the prompt into your agent chat (Cursor Composer, Claude Code, or Andromity). The agent automatically configures the skill instruction file into your workspace.
Yes, LangGraph Multi-Agent Workflows is 100% free and open-source under the mit license on AgenticMarket with no per-call fees or paywalls.
Replace with description of the skill and when Claude should use it.
Build stateful AI agents, durable workflows, real-time WebSocket communication, and MCP servers on Workers.