Planton Skills

Agent skills that let your coding agent work with Planton: compose cloud infrastructure, validate it, deploy it, and run services -- from Cursor, Claude Code, Codex, Gemini CLI, and any other host that reads the Agent Skills format.

Ask your agent for "a Postgres for this service in dev" and it writes a grounded manifest under infrastructure/ in your repository, validates it with the planton CLI, tells you what it will cost, and asks before applying.

What is here

Skill	What it carries
skills/planton	The working craft: Infra Charts and manifest sets, the compile loop, wiring resources by reference, deployed projects, service registration, push-to-deploy, CI/CD, and the boundaries that never bend (no mutation without consent, never outside your repository).
skills/multi-cloud-catalog	The component reference pack, shipped inside the skill: one page per cloud component across every supported provider (spec fields, validation rules, outputs, wiring), the catalog-wide reference graph, and verified fact sheets for cost, security posture, and runner permissions. Facts are read from these files at answer time, never recalled from memory.

Install

One command, any agent (the skills CLI detects your agents and offers project or global scope):

npx skills add plantonhq/skills

Target one agent explicitly, or install for every agent on the machine:

npx skills add plantonhq/skills --agent cursor
npx skills add plantonhq/skills --agent claude-code --global
npx skills add plantonhq/skills --agent '*' -y

Claude Code marketplace:

/plugin marketplace add plantonhq/skills
/plugin install planton@planton

Manual (any agent): clone this repository and copy or symlink skills/planton and skills/multi-cloud-catalog into your agent's skills directory -- ~/.cursor/skills/, ~/.claude/skills/, ~/.codex/skills/, or .agents/skills/ inside a project. SKILL.md must sit directly inside each skill folder.

Pair it with the CLI

The skill does its best work with the planton CLI on your PATH: validation and schema lookups run fully offline, and one planton login unlocks the compile loop, lookups, and deploys against your organization.

brew install plantonhq/tap/planton   # macOS; other platforms: https://planton.ai/docs/cli
planton login

Optional: the platform's tools over MCP

Without the CLI, or in addition to it, your agent can reach the platform's own operations (build, apply, deploy, read your organization's estate) through the hosted Planton MCP server. Create an API key in the Planton console, export it as PLANTON_API_KEY, then:

• Claude Code: /plugin install planton-platform-tools@planton

• Cursor (~/.cursor/mcp.json or .cursor/mcp.json):

  {
    "mcpServers": {
      "planton": {
        "type": "http",
        "url": "https://mcp.planton.ai/",
        "headers": { "Authorization": "Bearer ${env:PLANTON_API_KEY}" }
      }
    }
  }

The full guide, including the first journey end to end, is at planton.ai/docs/coding-agents.

Versions

Every commit on main is a Planton release, tagged with the release's tag (for example v0.5.41). The content under skills/ is byte-identical to the checksummed skill archives that release published to Planton's downloads CDN, and it is what Planton Desktop and the hosted assistant run at that version. npx skills update brings the latest release; pin a release with the tag.

Where the content comes from

skills/ is release output. The skills are authored in plantonhq/planton under skills/, linted against the Agent Skills specification on every pull request, and copied here by that repository's release lane. Pull requests that change files under skills/ in this repository are refused with a pointer to the source; improvements are welcome there.

License

Apache-2.0, the same as the source repository.