Nov 17, live webinar: ECS to EKS without freezing the roadmap for a quarter

A control plane for your agents

Qovery runs your apps and deployments. The Agents tab does the same for the coding agents that work on them: what they know, what they can do, and what they did

Free for 3 users and unlimited agents on your own cluster. See how every agent PR gets an environment

Qovery Agents interface

Trusted by 200+ companies to simplify their infrastructure burden

GetsafeSpikoCharles.coRentalReadyHyperlineThe Mobile First CompanyTalkspaceAlanPowensProsperityPivotZiplineTomorro

How it works

The future of infrastructure is agentic

Agents can do real infrastructure work on their own when every operation is checked at the API. Roles were made for humans. An agent should get a policy that allows exactly its job.

  • Every agent goes through the Qovery API, from the Console, the CLI, Terraform or MCP.
  • Custom skill presets set what your agents read before they act.
  • API Policy Tokens set what each call may do. Anything else is refused before it runs.
  • Every call lands in the audit log, under the member’s name or the token’s.
Read the talk: who gives the agent production access?
Six coding agents reach the Qovery API through the Console, the CLI, Terraform or MCP. The agent control layer holds skill presets, API Policy Tokens with a policy that allows reading logs, deploying and deleting old previews but refuses deleting the production database, activity and audit logs. Below it sit Qovery environments, Kubernetes and your own cloud.
Custom skill preset editor: each Qovery skill used as is, with team instructions added, replaced or disabled, and the team instructions for qovery-deploy

Guide: your own version of the Qovery skills

A custom skill preset sets which Qovery skills your members’ agents load, and what your team adds on top. Make one per team or per person, or start from a template for their profile. Agents read a short list first and load only the skill a task needs.

  • Start from the full Qovery skills, a lean CLI-first skill, or none
  • Use each Qovery skill as is, add instructions, replace it or turn it off
  • Add your team’s rules and your own skills
  • Assign it to the organization, a role or one person
  • Every save is checked by rules and a model review. Members see what their agent receives
How skill presets work
policy.regoAPI Policy Token
# Read everything. Deploy one application. Nothing else.default allow := false allow if read_only_accessallow if deploy_access app := "5bf547f9-b445-44bc-aab8-7f341187d362"read_only_access if input.request.method in {"GET", "HEAD"}deploy_access if input.request.path == ["api", "application", app, "deploy"]
  • GET /api/environment/…/applicationallowed
  • POST /api/application/5bf547f9…/deployallowed
  • DELETE /api/environment/…401 refused

Control: a policy for each agent, not a role

An API Policy Token carries a policy written in Rego instead of a role. Qovery checks it on every API request, from the Console, the CLI, Terraform and MCP. Allowed actions run through the same pipeline as a human click. Everything else is refused.

  • Deny by default: a request runs only if a rule allows it
  • Scope a rule to one environment, one service or one action
  • Every call is in the audit log under the token’s own name
  • Describe the job in plain English and the qovery-policy-token skill writes and tests the policy
How API Policy Tokens work
Agents Activity page with filters, skill loads, allowed and denied actions, and a chart of actions over 30 days

Observe: what agents did, and what was refused

Activity shows what your members’ agents did through the skills and the MCP server, and what each policy token did, allowed and denied, in one view. Only actions taken through Qovery are visible. Prompts and code are never collected.

  • Filter by member, policy token, skill, source and outcome
  • See every refused call in one list
  • Follow which skill led to which action
  • Share a filtered view with a link
How Activity works
Automate catalog of recurring agent jobs by level, each with its suggested trigger

Automate: agents that run on their own

Pick a recurring job and give it a policy token that allows only that job. It runs on a schedule or on an event, as an Agent Task on Qovery or from the coding agent you already use: Claude Code, Codex, Cursor, OpenCode, Gemini CLI or VS Code Copilot. Start with the safest jobs, then move up.

  • Investigate an incident, or verify a fix in a preview
  • Stop idle environments on weeknights and clean up old previews
  • Right-size applications every week
  • Roll back when errors rise after a deploy, and patch security issues every week
See the agent policy examples
Build optimization analysis with review confidence, estimated time saved, a summary and the improvements to make

Optimize builds: an agent with one job

The Build Optimizer reads the latest Docker build of a service and finds why it is slow, most often layers that miss the cache. It returns the fixes, a reworked Dockerfile when one applies, and a prompt for your coding agent.

  • Reads the Dockerfile, the build logs and the cache history
  • Never changes your Dockerfile, build settings or image
  • Never starts a build: you review and apply the fixes
How the Build Optimizer works

Connect each machine once

Install the Qovery skills, the CLI or the MCP server on a laptop or a CI runner. The agent then loads your organization’s preset. Anything that speaks MCP can operate Qovery.

AI, developer and collaboration tools that can operate Qovery through MCP

Free plan

Free for your coding agents

Every pull request your agent opens gets its own live environment on your Kubernetes cluster. Free for 3 users and unlimited agents, with 1,000 deployment minutes a month. No time limit, no credit card.

  1. 01

    Connect your Kubernetes cluster

    Kubernetes 1.31 or later: EKS, GKE, AKS, k3s or on-premise. Qovery installs on the cluster you already run, and your workloads stay in your account.

  2. 02

    Install the Qovery skill or MCP in your agent

    One command for Claude Code, Cursor, Codex or any MCP client. Each agent gets its own token, scoped by your roles.

  3. 03

    Every agent PR gets a live environment

    Turn on preview environments once. Each pull request your agent opens gets its own copy of the environment, with its own URL, and it is deleted when the PR closes.

Built for platform teams

What holds when agents work on your infrastructure

Deny by default

A policy token can do nothing until one of its rules allows it.

One check for every interface

The same policy applies to calls from the Console, the CLI, Terraform and MCP.

Checked before members get it

Rule checks and a model review run on every preset save. Nothing is saved until both pass.

Nothing injected silently

Agents tell members which preset and skills they loaded, and members can read all of it.

Full history

Each preset edit is a revision you can compare and restore. Each skill load is in the audit logs.

No prompts, no code

Activity records actions taken through Qovery. It never collects prompts or code.

Ready to build the future of your infrastructure?