Swytchcode Documentation
Build AI agents that safely discover and execute real APIs with Swytchcode. Explore guides, references, and examples to get started.
Start here
Setup and onboarding
Follow the setup flow, install the CLI, and move from first run to production-ready integrations.
- Getting started guide
- Account setup
- Plugin installation
Pick a framework
Jump straight into the framework-specific examples and launch a working integration quickly.
- OpenAI SDK
- Anthropic SDK
- LangGraph and more
Browse practical patterns
See real-world agent workflows, API integrations, and production recipes you can adapt immediately.
- Cookbook examples
- Agent workflows
- Integration patterns
Learn the platform
What is Swytchcode
Understand the execution layer that sits between your AI agent and production APIs.
Cookbook examples
Explore recipe-style use cases and implementation patterns that map directly to real deployments.
Runtime SDK
Use the runtime APIs and SDK surfaces to build, test, and run integrations in production.
Configuration
Manage environment variables, provider setup, and runtime configuration for secure execution.
CLI reference
Use the CLI to execute calls, manage auth, and operate agent workflows from the terminal.
MCP reference
Look up MCP-specific commands and tools exposed by the Swytchcode CLI.
Framework quickstarts
OpenAI SDK
Get started with the OpenAI SDK integration quickstart.
Anthropic SDK
Build with Claude and agent-ready models using Swytchcode.
Vercel AI SDK
Follow the Vercel integration quickstart for deployable AI workflows.
LangGraph
Connect Swytchcode to LangGraph workflows and agents.
CrewAI
Use Swytchcode with CrewAI for real-time agent execution.
Hermes
Launch Hermes-powered agents with Swytchcode integration.
OpenClaw
Explore the OpenClaw quickstart for API-first agent workflows.