Sealos Skills plugin
An agent plugin: the phrase "deploy X to Sealos" runs sign-in, deploy path choice, databases, HTTPS and post-deploy verification.
- 84 ★ on the repository
- Last commit
Broad access
Why this level
- The skill creates and deletes apps, databases and storage in your Sealos cloud account
- Sign-in uses a kubeconfig with cluster rights stored in `~/.sealos/kubeconfig`
- The build path pushes an image to a registry
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add plugins/sealos-skillsDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- The skill creates and deletes apps, databases and storage in your Sealos cloud account
- Sign-in uses a kubeconfig with cluster rights stored in `~/.sealos/kubeconfig`
- The build path pushes an image to a registry
This is third-party code. Review the repository files before installing.
What it does
The plugin installs the `use-sealos` skill for Sealos Cloud into the agent. It takes a deploy request, checks credentials with `sealos-api.py status` and signs in through OAuth2 device flow if needed, or takes a kubeconfig from the web console into `~/.sealos/kubeconfig`. The skill picks one of three paths: a store template (200+ templates in about 3 minutes), an official Docker image with a generated template for a self-hosted app (5 to 10 minutes), or a source build through a Dockerfile, `docker buildx` for `linux/amd64` and a registry push. Every path goes through the Template API, so a deployment shows in the Sealos UI as one instance and is removed as a unit. After the rollout `wait-app.sh` checks readiness and the URL. Reference files cover KubeBlocks databases, storage, logs and debugging.
Who it is for. Developers deploying apps and databases to Sealos Cloud who want to hand the routine to an agent in Claude Code, Codex or Cursor.
Good fit when
- A self-hosted app such as a popular wiki or tracker has to come up with a database and public HTTPS
- Your own project with a Dockerfile must be built for `linux/amd64`, pushed to a registry and deployed
- An app from the template store needs the fastest path in a couple of minutes
- After a rollout you need to check status, read logs and find the cause of a crash
Not a fit when
- The deployment is not going to Sealos Cloud: the skill knows only its Template API and manifests
- You need separate Compose or Dockerfile converters, a desktop builder and a cloud readiness check: version 2.0 moved them to an archive branch
- You have no Sealos account: sign-in uses OAuth2 or your kubeconfig
Example request
Deploy Plausible Analytics to Sealos with a database and public HTTPS, then check that the URL responds.Limitations
`kubectl` and `python3` (standard library only) are required, plus `docker` and `gh` for the build path. A Sealos account is needed. The kubeconfig is saved to `~/.sealos/kubeconfig` and the host may need access to that folder. The `sealos-deploy` and `k8s-kaniko-job` skills target the managed Brain sandbox and do nothing outside it. In Gemini CLI and Qwen Code the plugin acts as a context extension without slash commands. Version 2.0 replaced the previous eight skills with a single `use-sealos`.
How to disable. In Claude Code run `/plugin uninstall sealos@sealos-skills` and remove the marketplace. For Codex use plugin management, for Cursor delete the `~/.cursor/skills/use-sealos` link. Created apps are removed in the Sealos UI as instances, and `~/.sealos/kubeconfig` should be deleted separately.
FAQ
How do I install the plugin in Claude Code?
Run `claude plugin marketplace add labring/sealos-skills`, then `claude plugin install sealos@sealos-skills` and `/reload-plugins`. Interactive work goes through `/sealos`.
How do I install in Codex or Cursor?
In Codex: `codex plugin marketplace add labring/sealos-skills` and `codex plugin add sealos@sealos`, invoked as `$sealos`. In Cursor paste the repository URL under Settings, Plugins. The universal route is `npx skills add labring/sealos-skills`.
How does the skill sign in?
Through OAuth2 device flow. Or copy a kubeconfig from the Sealos web console into `~/.sealos/kubeconfig` and the skill will use it.
Which deploy path should I pick?
The skill picks itself: a store template first, then an official image with a generated template, and a source build for your own code. All three are tracked as an instance in the Sealos UI.
What do sealos-deploy and k8s-kaniko-job do?
They are skills for the managed Brain sandbox: GitHub deploy with MCP signals and an image build through a Kaniko Job. Outside such a sandbox they perform no actions.
Discussion
No questions or reviews yet.
Related
PluginsPlugin with one skill: the agent prepares a project, deploys a container to the Cloudish cloud and returns the live URL.
KubeShark skill: the agent finds the failure mode first, then writes manifests, Helm and Kustomize to official practice with a rollback plan
A CLI and Kubernetes operator that runs Claude Code, Codex, Gemini, OpenCode and Cursor as tasks and sessions inside a cluster
Docker Skills
docker-skills
Official Docker skills for agents: Dockerfile and builds, Compose, Sandboxes, Docker Agent and a skill that confirms irreversible actions
Cloudflare's official agent skills: Workers, Durable Objects, wrangler, Pages, storage and an MCP server
A CLI for local and cloud Kubernetes clusters with GitOps and secrets; ships an MCP server and a plugin for Claude Code
README in short
The page describes a plugin for Sealos Cloud in the Agent Plugins format and as a DeepSeek Harness profile. It shows the repository tree, installs for Cursor, Claude Code, Codex, Qoder, DeepSeek Harness, OpenClaw, CodeBuddy, Gemini CLI, Qwen Code and skills.sh, the first run, the three-path deploy strategy and requirements. It ends by explaining that version 2.0 replaced eight skills with one.