Getting started

Running in 15 minutes

This guide takes you from a clean checkout to a working webhook API, background worker, and MCP server you can point Cursor or Claude Desktop at.

Prerequisites

Python 3.9+
kubectl for Kubernetes features
Cloud CLI tools (optional): aws, gcloud, az

1. Install

# Clone and enter the repo

$ git clone https://github.com/open-devops-agent/open-devops-agent.git

$ cd open-devops-agent

# Create a virtual environment

$ python -m venv venv && source venv/bin/activate

# Core dependencies

$ pip install -r requirements.txt

# Cloud providers (optional)

$ pip install boto3 google-cloud-compute azure-mgmt-compute

2. Configure

Copy the example environment file and fill in the minimum required variables. Orgs sharing one deployment can instead register keys per-org via the API or MCP env.

$ cp .env.example .env

# Required for the built-in agent loop

ANTHROPIC_API_KEY=your-claude-api-key

# Critical — alerts before dangerous operations

EMAIL_ENABLED=true

EMAIL_SMTP_HOST=smtp.gmail.com

EMAIL_TO=sre-team@yourcompany.com

# Safety — keep false in production until reviewed

AUTO_APPLY=false

3. Run locally

# Webhook API + background queue worker

$ devops-agent serve

# MCP server for Cursor / Claude Desktop / custom agents

$ devops-agent mcp

4. Test with an incident

$ curl -X POST http://localhost:8000/webhook/manual \

-H "Content-Type: application/json" \

-H "X-Org-ID: acme-corp" \

-d '{ "type": "k8s", "namespace": "default", "pod": "myapp-pod", "description": "CrashLoopBackOff test" }'

# wait ~30s, then check the org-scoped audit log

$ curl "http://localhost:8000/audit?org_id=acme-corp"

5. Deploy to Kubernetes

$ kubectl create namespace devops-agent

$ kubectl create secret generic agent-secrets \

--from-literal=ANTHROPIC_API_KEY=your-key \

--from-literal=GITHUB_TOKEN=your-token \

-n devops-agent

$ kubectl apply -f k8s/ -n devops-agent

Next steps

→ Full API reference → Documentation index → Read the security guarantees before production