3.1 KiB
3.1 KiB
description
| description |
|---|
| Connect the Conductor CLI, SDKs, and agent runtime to Developer Edition or a local server. |
Connect to Conductor
Recommended: Orkes Developer Edition
Create a free account, application, and access key in Orkes Developer Edition. Then set the following environment variables.
export CONDUCTOR_SERVER_URL=https://developer.orkescloud.com/api
export CONDUCTOR_AUTH_KEY=<your-access-key>
export CONDUCTOR_AUTH_SECRET=<your-access-secret>
You can then proceed to configure the local CLI and core SDKs.
Install the CLI
The CLI registers workflows and starts executions against your chosen Conductor server.
npm install -g @conductor-oss/conductor-cli
Local server alternative
Use when you need a self-managed development server. It requires Java 21+ and Node.js.
conductor server start
export CONDUCTOR_SERVER_URL=http://localhost:8080/api
conductor workflow list
AI and agent credentials
Configure model access and credentials for AI workflows and agents.
-
Developer Edition: add an integration for your model provider under Integrations
-
Local server: export the provider key before starting the server so it inherits it. For example:
export OPENAI_API_KEY=<your-openai-api-key> conductor server start
Docker
You can also run Conductor via the official Docker container.
docker run --rm -p 8080:8080 conductoross/conductor:latest
export CONDUCTOR_SERVER_URL=http://localhost:8080/api
conductor workflow list
Next steps
Once you have Conductor up and reachable, choose what you want to build.
Your first workflow & worker
Author and run a durable workflow in your chosen language.
Run your first agent
Author and run a Conductor Agent with Python, Java, TypeScript/JavaScript, or C#.
Bring a framework agent
Run an existing OpenAI Agents, LangChain, LangGraph, or Google ADK agent through Conductor.
No-code
Register and run a workflow with the CLI and JSON