Qovery Skills is a collection of AI agent skills officially provided by Qovery, used to deploy and troubleshoot applications on Kubernetes. It follows the Agent Skills open standard and can be used with OpenCode, Claude Code, and any agent that supports SKILL.md, and is compatible with more than 30 AI programming tools.
Purpose
This skill package enables AI programming agents to directly operate the Qovery platform, completing a series of tasks from onboarding, application deployment, troubleshooting, cost optimization, and deployment acceleration, to preview environment creation, Terraform conversion, policy token generation, and organization assessment.
Main Capabilities
The skill package contains eleven skills in total:
- qovery: A general entry point, used to list environments, stop services, check deployment status, and more.
- qovery-onboard: Acts as a personal cloud architect, understands the user's role, existing resources, and needs, recommends cloud providers, regions, cluster types, environment structure, secure defaults, and RBAC roles, and performs the complete setup. Supports BYOK (bring your own Kubernetes) and migration from platforms such as Heroku, Vercel/Netlify, and Render/Railway.
- qovery-deploy: Analyzes the codebase, detects language, framework, ports, database requirements, and environment variables; generates production-grade templates for missing Dockerfiles (supports more than 12 frameworks); deploys via CLI+API or Terraform; configures databases (container mode for development/testing, managed mode such as AWS RDS for production); handles Helm charts, Terraform modules, lifecycle jobs, and scheduled tasks; monitors deployments and automatically fixes build errors, port mismatches, health check failures, missing environment variables, and OOM issues.
- qovery-troubleshoot: Integrates the Qovery MCP Server for diagnostics, runs 8 layers of systematic diagnosis, matches more than 20 error patterns, includes 10 preset playbooks, automatically fixes Qovery configuration (asks before modifying user code), and generates runbooks.
- qovery-optimize: Collects business context, analyzes historical resource consumption, optimizes across 7 dimensions, estimates external cloud resource costs, generates detailed cost reports (Markdown and CSV), applies changes via the Qovery API or Terraform, considers seasonal patterns, and provides Kubecost deployment recommendations.
- qovery-speedup: Measures the complete deployment pipeline, breaks down the time consumed by substeps, generates build runner usage reports, classifies bottlenecks by ownership, and proposes recommendations for Dockerfile optimization, health check tuning, deployment phase parallelization, and image pull optimization.
- qovery-preview: Creates a preview environment for the current branch or PR.
- qovery-terraform: Converts existing Qovery configuration into Terraform infrastructure as code.
- qovery-policy-token: Creates restricted API tokens, for example, ones that can only deploy but not delete, or that are limited to specific environments or services.
- qovery-assess: Read-only assessment of organization configuration, runs 147 checks, covering twelve categories: compliance, clusters, topology, reliability, anti-patterns, security, variables and secrets, delivery, logs, change sources, cost, and disaster recovery, and generates shareable reports and findings CSV.
- qovery-signup: Helps new users sign up for Qovery.
Usage
Installation command (globally install all eleven skills):
curl -fsSL https://...
Install only in the current project:
...
The installer automatically places all eleven skills (including the reference/, templates/, and examples/ subdirectories) in the correct locations so that any compatible tool can discover them. Updating to the latest version only requires re-running the install command, which will overwrite the old version.
After installation, trigger skills through natural language prompts, for example:
- "I'm new to Qovery, help me get started" (triggers qovery-onboard)
- "Deploy my application with Qovery" (triggers qovery-deploy)
- "My deployment is failing" (triggers qovery-troubleshoot)
- "Optimize my Qovery costs" (triggers qovery-optimize)
- "My deployments are slow" (triggers qovery-speedup)
- "Create a preview environment for the current branch" (triggers qovery-preview)
- "Terraformize my Qovery setup" (triggers qovery-terraform)
- "Create a restricted API token that can only deploy, never delete" (triggers qovery-policy-token)
- "Assess / audit / review my Qovery organization" (triggers qovery-assess)
Slash commands are also supported, such as /qovery, /qovery-preview, /qovery-terraform, /qovery-policy-token, /qovery-assess.
Applicable Scenarios
Suitable for development teams, DevOps engineers, CTOs, and founders who use the Qovery platform to deploy and manage applications on Kubernetes. Especially suitable for scenarios that require quickly getting started with Qovery, migrating from other platforms, automating deployment processes, troubleshooting deployment failures, optimizing cloud costs, speeding up deployments, creating previewenvironments, converting configurations to Terraform, generating restricted tokens, or conducting organizational security and compliance assessments.
Limitations
The skill package itself is an open-source repository, and the complete URL for the specific install command is truncated in the provided materials, so the full content is not currently available. Some features (such as cost estimation and resource consumption analysis) depend on cluster observability, and suggestions will be given if it is not enabled. The assessment skill is read-only and will not modify any configuration. Specific charging, pricing, and platform support details are not currently provided.
