examples

# Dokploy Examples Step-by-step walkthroughs covering the most common Dokploy workflows. Each scenario uses MCP tools as the primary method, with equivalent curl commands shown alongside where relevant. --- ## Scenario 1: Deploy a Web Application Deploy a web app from a GitHub repository to a running application with a custom domain and HTTPS. Covers: project creation, application setup, GitHub integration, environment variables, domain configuration, and deployment. See full walkthrough: [references/deploy-app.md](references/deploy-app.md) --- ## Scenario 2: Provision a Database Create, configure, and expose a PostgreSQL database with external access and automated backups. Covers: database creation, deployment, external port configuration, connection string construction, and backup scheduling. See full walkthrough: [references/provision-database.md](references/provision-database.md) --- ## Scenario 3: Deploy a Docker Compose Stack Deploy a multi-container application using a docker-compose.yml file with service-level domain routing. Covers: compose service creation, compose file content, environment variables, deployment, and per-service domain assignment. See full walkthrough: [references/compose-stack.md](references/compose-stack.md) --- ## Scenario 4: Debug a Failed Deployment Diagnose and recover from a failed deployment end-to-end — locate the failed run, read the build log, inspect the container, check Traefik, optionally AI-summarise, apply the fix, and verify. Covers: `deployment-all` filtering, reading runtime + build logs over MCP (`application-readLogs` / per-container `compose-readLogs` / `deployment-readLogs`, v0.29.0+ (current v0.30.7)), `docker-getContainersByAppLabel` + `docker-getConfig` introspection (with env values redacted by default), v0.30 host diagnostics (`docker-getServerHealth`), `ai-analyzeLogs { aiId, logs, context }` integration, and choosing the right recovery action. See full walkthrough: [references/debug-failed-deploy.md](references/debug-failed-deploy.md)