Penpot AI Kit plugin

Skills and brief templates for an agent working in a Penpot file over MCP: components, tokens, accessibility, change previews.

  • 121 ★ on the repository
  • Last commit
  • CC-BY-4.0
Plugin

Extended access

Why this level

  • An agent gets write access to a Penpot file via the MCP key; changes apply after preview and confirmation.
All reasons and checks

penpot/penpot-ai-kit

Install

In your terminal, with SkillFoxx CLI

npx skillfoxx add plugins/penpot-ai-kit

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • An agent gets write access to a Penpot file via the MCP key; changes apply after preview and confirmation.

This is third-party code. Review the repository files before installing.

What it does

Penpot AI Kit adds ready skills and instructions to the Penpot MCP so an agent works in a real file more reliably than from a one-off prompt. The agent uses your components and design tokens, follows system rules, checks accessibility and shows changes before applying them. Brief templates cover typical tasks: `/penpot-ai-kit:design-brief`, `/penpot-ai-kit:deck-brief` and others. The kit installs with a simple phrase: the agent reads `INSTALL.md`, detects the client, asks about remote or local Penpot and the MCP key, connects the MCP, wires skills and verifies the live connection. The installer copies the kit to `~/.penpot-ai-kit`, keeps an install manifest and never prints the key. A Claude Code plugin exists too; skills are named `penpot-ai-kit:penpot-*`. Claude Code, Claude Desktop, Cursor, Windsurf, OpenCode and Codex are supported.

Who it is for. Designers and design-system teams in Penpot who want an agent to edit layouts following their rules.

Good fit when

  • An agent must assemble a screen from existing components and tokens rather than draw from scratch.
  • Changes to a file should be previewed before they are applied.
  • A layout is checked for accessibility and design-system compliance.
  • A design or deck brief is filled from a template.

Not a fit when

  • There is no Penpot MCP key or MCP access is closed.
  • Layouts live in another tool, not in Penpot.
  • The agent must not get write access to the working file even with confirmation.

Example request

Build a sign-in screen from our design system components, show the changes before applying them and check text contrast.

Limitations

The kit needs the Penpot MCP and an access key; without them the skills do not work. Do not combine the plugin and the Node.js installer for Claude Code or skills would fire twice: if you used the installer, remove it first with `node scripts/install/uninstall.mjs --yes`. The clone is only a source, and the installer copies the kit to `~/.penpot-ai-kit`. The CC BY 4.0 license allows use with attribution. Rules and templates rely on your tokens and components, so results depend on how orderly your Penpot file is. The last commit dates from October 2026.

How to disable. Tell the agent "Uninstall the Penpot AI Kit" or run `node scripts/install/uninstall.mjs --yes`: skill wiring, the `penpot` MCP entry and the installed copy are removed. For the plugin run `/plugin uninstall`.

FAQ

Is a separate Penpot MCP needed?

Yes. The kit relies on the Penpot MCP and does not work without it. Setup is described in `docs/setup-remote.md` and done by the installer.

How do I update the kit?

Tell the agent "Update this Penpot AI Kit": one `update.mjs` step refreshes the installed copy and re-wires skills in every client.

Can I use both the plugin and the installer?

Not recommended for Claude Code: skills would trigger twice. Pick one method.

Which brief templates exist?

Among them `design-brief` and `deck-brief`; in the plugin they appear as commands such as `/penpot-ai-kit:design-brief`.

What does the installer do with the key?

It never shows or prints it. It asks confirmation before every change and is safe to re-run.

Discussion

No questions or reviews yet.

Plugins

The official Pipefy toolkit for agents: an MCP server, a CLI and skills that work with Pipefy workflows as your signed-in account.

PluginBroad access53Repository stars
Official

A Claude Code plugin and CopilotKit skills: the agent reads current docs and source, checks a project with verify and sets up Slack channels

PluginExtended access38.7KRepository stars

Potpie builds a context graph from code, history and team decisions and wires it into Codex or Claude Code through a skill bundle

CLIExtended access5.8KRepository stars

An MCP builder agent that picks servers from the official registry and composes them into your own agentic MCP server.

MCP serverExtended access105Repository stars

A mirror of Lauren Tan's pstack set from Cursor: over 50 skills, the poteto-mode with 23 playbooks and 24 principles for strict agent work on code

SkillBroad access4.1KRepository stars
Official

A local MCP gateway: set servers up once and use them in Claude Code, Codex, Cursor and other agents, with keys in the OS keychain

MCP serverBroad access239Repository stars

README in short

The README gives a fast path through a phrase for the agent, a list of follow-up requests, plugin install and notes on limits. It warns about plugin and installer compatibility.

README badge

Report a problem