Elestio CLI & Agent Skill: Let AI Agents Deploy and Manage Your Infrastructure
Two months ago, we started asking ourselves a simple question: what if your AI coding assistant could deploy infrastructure the same way it writes code? Not through a dashboard. Not by generating Terraform files you have to review and apply. Just... deploy it. Say "I need a PostgreSQL database" and have a production-ready, fully managed instance running two minutes later.
So we built it. Today we're open-sourcing two things: the Elestio CLI, a full command-line tool for managing the Elestio platform, and the Elestio Skill, which teaches AI agents how to use it. Deploy any of our 400+ open-source services, push your own code via CI/CD, manage backups, SSL, firewall, and monitoring. All from a single conversation.
Install It in 10 Seconds
One command installs both the CLI and the agent skill:
curl -fsSL https://raw.githubusercontent.com/elestio/elestio-skill/main/scripts/install.sh | bash
Or via skills.sh:
npx skills add elestio/elestio-skill
The skill follows the Agent Skills open format and supports Claude Code, OpenAI Codex, OpenCode, and Cursor. Re-run to update.
If you just want the CLI without the agent skill:
npm install -g elestio
Then configure your credentials:
elestio login --email "you@company.com" --token "your-api-token"
elestio auth test
# → [SUCCESS] Authenticated as you@company.com
You can grab your API token from the Elestio dashboard. That's it. Your agent can now deploy infrastructure.
What the Skill Actually Does
The installer does two things: it installs the Elestio CLI globally (npm install -g elestio), and it drops a SKILL.md file into your agent's skills directory. That file teaches the agent every command, every workflow, and every edge case. The CLI itself wraps the entire Elestio API into clean, composable commands.
Here's what that looks like in practice:
# Search the catalog
elestio templates search postgresql
# Deploy it
elestio deploy postgresql --project 112 --name prod-db
# Get your credentials when it's ready
elestio credentials <vmID>
# → URL, admin user, password
That's it. The agent handles provider selection, region, sizing, and waits for deployment. Your database shows up fully configured with automated SSL, backups, and monitoring.
But the catalog is just the beginning. The skill also handles custom deployments:
# Deploy your app from GitHub (fully automated)
elestio deploy cicd --project 112 --name my-cicd
elestio cicd create --auto \
--target <vmID> --repo your-org/your-app \
--mode github --name my-frontend
The CLI automatically discovers your Git account, finds the repo, generates an optimized multi-stage Dockerfile, builds the image, starts the container, and verifies HTTP 200. The whole pipeline creation takes about two minutes.
Not Just Deploy: Full Lifecycle Management
We didn't build a deployment button. We built a full infrastructure management layer:
| Capability | What It Does |
|---|---|
| 400+ Templates | One-command deploy for PostgreSQL, Redis, WordPress, Grafana, n8n, and hundreds more |
| Custom CI/CD | GitHub, GitLab, Docker Compose pipelines with auto-Dockerfile generation |
| 9 Cloud Providers | Netcup, Hetzner, AWS, Azure, GCP, DigitalOcean, Vultr, Scaleway, Linode |
| Backups | Local, remote, and S3-compatible external backups |
| SSL & Domains | Auto-SSL via Let's Encrypt, custom domain binding |
| Firewall | Programmatic firewall rules per service |
| Auto-Updates | Scheduled OS and app updates with security-only options |
| Snapshots & Volumes | Provider-level snapshots and NVMe block storage management |
| Billing | Real-time cost tracking per project and service |
Every Elestio service runs on a dedicated VM (not shared Kubernetes), across 9 cloud providers and 100+ regions. The CLI lets agents pick the optimal configuration for each workload.
Why This Matters
The AI agent ecosystem is moving fast. Developers are already using agents to write code, run tests, and debug issues. But when it comes time to deploy, they switch back to dashboards and CLIs designed for humans.
That context switch is expensive. Your agent understands the full project, knows which services are needed, and can reason about dependencies. Why force it to stop at the infrastructure boundary?
With the Elestio Skill, the agent goes from "your Next.js app needs a PostgreSQL database and a Redis cache" to three running services with credentials, SSL, and backups configured. No tickets, no clicking through cloud consoles, no YAML files.
What's Next
We're actively developing both the CLI and the skill based on real-world agent workflows. Multi-service orchestration (deploy a full stack in one conversation), environment cloning, and cost optimization recommendations are all on the roadmap.
Infrastructure starts at $14/month in EU ($16 in US) for a dedicated VM with 2 CPU, 4 GB RAM, and 60 GB NVMe storage on Netcup. Every service includes automated SSL, daily backups, OS security patches, and 24/7 monitoring.
Both projects are open-source under MIT license: the CLI (also on npm) and the Skill. Star them, fork them, extend them. Check out the full catalog of 400+ services or create your Elestio account to get started.
Thanks for reading! See you in the next one.