Open supply initiatives often start with a easy perception: software program turns into extra succesful, extra resilient, and extra significant when it’s created within the open. For Penpot, this perception will not be a slogan. It’s the basis of the whole product. Openness guides how options are imagined, how selections are reached, how the neighborhood participates, and extra lately, how Penpot approaches AI.
As AI continues to reshape inventive work, many design instruments have launched clever options developed behind closed doorways. Penpot is selecting one other means. As an alternative of presenting a completed thought of what AI ought to do, the staff invitations the neighborhood to discover what AI might do. A lot of what makes a design significant isn’t written instantly right into a file.
As Àlvaro Tejero famous at Penpot Fest, a design carries intent and relationships that aren’t recorded within the code, but designers acknowledge them immediately.
The brand new MCP Server sits on the coronary heart of this mindset. It isn’t a product announcement however an invite to discover collectively what AI assisted design might grow to be.
Why Open Collaboration Issues for Design Instruments
The design trade strikes rapidly. New practices and applied sciences seem at a gentle tempo, and the boundary between design and improvement turns into thinner yearly. On this surroundings, a instrument formed by a small closed staff can start to really feel inflexible or disconnected from actual every day work.
Penpot follows a unique philosophy. Openness will not be solely about releasing supply code. It’s about creating room for significant contribution. Clear processes and extensible instruments stop concepts from getting trapped inside a single firm. As an alternative, concepts enhance by means of actual world suggestions and experimentation. That is very true in design, the place which means comes from context, emotion, and intent as a lot as from seen parts. As Tejero noticed, individuals naturally perceive extra from a design than what’s written in knowledge or coordinates.
This surroundings permits contributors to affect the course of the product moderately than solely requesting options. That stage of collaboration is troublesome to match in proprietary ecosystems, and it turns into much more helpful when AI enters the image. The tempo of AI improvement is just too quick for a closed roadmap to anticipate every thing. A neighborhood can think about way more.
How Group Collaboration Shapes Penpot
Anybody who has adopted Penpot in recent times has seen the neighborhood form the instrument in seen methods. Giant options corresponding to Design Tokens and Variants have been influenced and sometimes stress examined by designers and builders lengthy earlier than they reached a secure launch.
The beta testing cycles illustrate this effectively. Every main characteristic is shared early with the neighborhood, who attempt it in actual initiatives, break it in sudden methods, and supply suggestions that helps refine the expertise lengthy earlier than remaining launch. Selections are made publicly and mentioned with openness, at all times grounded in actual workflows.
The identical collaborative spirit seems within the API ecosystem. By exposing the Plugin API and treating it as a firstclass extension level, Penpot encourages experimentation. Builders have already created workflow automation, integrations with code instruments, accessibility helpers, and early AI brokers that examine designs by means of the plugin layer. These experiments inform the product and assist the staff perceive rising wants.
This tradition of transparency and participation makes the subsequent step doable.
The MCP Server: A New Layer for Co Creation
At Penpot Fest 2025, the staff offered an early preview of the MCP Server. It’s a new interface that enables AI brokers and instruments to work together with Penpot designs by means of a versatile and mannequin impartial protocol. As an alternative of counting on screenshots or non-public interfaces, an AI system can use the identical Plugin API obtainable to people, solely with quicker entry to Penpot’s structured design mannequin.
This opens a door to workflows that weren’t doable earlier than.
In neighborhood demos, AI brokers might analyze design programs, extract model guidelines, refactor naming, generate semantic HTML and CSS, replace kinds from documentation, construct Storybook setups, and even flip tough sketches into extra structured parts.
Penpot constructed the MCP Server round a easy thought: AI turns into extra dependable when it could perceive specific design construction. As Tejero famous, giant language fashions have limits, however they will work way more precisely after they obtain organized info. By exposing Penpot’s inner graph of relationships as a substitute of uncooked pixels, the server makes semantic understanding doable.
This preview is deliberately open ended. The MCP Server doesn’t dictate how AI ought to behave. It offers a basis that designers, builders, and toolmakers can discover and adapt. This is the reason it was launched early. As Tejero defined, the objective was to let individuals check it, problem it, and share how they wish to use it.
By sharing the server at this stage, Penpot is basically saying: Here’s a basis. Assist us think about what needs to be constructed on prime of it.
Why Penpot Is Asking for Use Instances
As a result of AI in design workflows continues to be new, nobody can totally predict which concepts will grow to be important. Some designers need assist cleansing layer buildings. Others need alignment between design and code. Others need refactoring, naming consistency, documentation syncing, or instruments that discover excessive stage concepts.
As an alternative of guessing, Penpot needs to listen to instantly from the individuals who will use these instruments. The staff already understands the foundational items the MCP Server wants, however they need to study which particular use instances excite the neighborhood most. Penpot is reaching the purpose the place the server will evolve from a versatile, generic interface into one thing formed by actual, use-case-driven improvement—and neighborhood enter is what is going to information that transition.
These questions are answered finest by means of collaboration with a neighborhood that already consists of designers, frontend builders, open supply contributors, educators, and inventive coders.
The MCP Server turns into a catalyst for that collaboration: a versatile basis that may develop in instructions no single roadmap can predict.
Potential Use Instances
Under are among the early MCP Server experiments proven within the Penpot demos. They illustrate how AI brokers can work instantly with Penpot’s structured design mannequin to automate actual duties throughout design and code workflows.
Design to code
An AI agent can learn the construction of a Penpot design — its parts, structure, and tokens — and switch that into clear, semantic HTML and CSS. This strikes past easy export: the agent understands the design’s relationships and produces code that displays them.
Prototype interactions
As an alternative of manually wiring up interactions, an AI agent can generate an interactive prototype from the design. It might probably take present belongings or HTML recordsdata and join them right into a functioning expertise, serving to groups discover habits earlier within the course of.
Scribble to design
A tough sketch can be utilized as the idea for an precise structured structure. The agent interprets the scribble’s visible intention and produces a extra refined design that follows the identical normal construction, giving designers a quicker solution to transfer from thought to draft.
Design System documentation + Design tokens
An agent can examine a design or design system, determine token utilization and construction, and generate documentation robotically. This makes it simpler to take care of a shared supply of fact and hold groups aligned as programs develop.
These experiments barely scratch the floor. They present how the MCP Server turns Penpot right into a playground for automation, creativity, and fully new workflows, the place AI doesn’t change the designer, however removes the friction between creativeness and execution.
Do you see the unbelievable potentialities but?
How You Can Get Concerned
When you care about the way forward for design, design programs, inventive tooling, or AI assisted workflows, Penpot welcomes your perspective. Whether or not you design advanced interfaces, take pleasure in automation, or experiment with AI brokers, your concepts might help information the expansion of the MCP Server.
You possibly can contribute by:
- Sharing your use instances and concepts
- Testing workflows when the beta opens
- Becoming a member of discussions within the Penpot neighborhood
- Exploring the Plugin API or early MCP examples
- Providing suggestions on workflows, naming, or design to code processes
To hitch the dialog round early experiments and use instances, you may go to the neighborhood dialogue house. (You’ll want to join free to entry the thread.)
Even small concepts can affect the course of the undertaking. A lot of Penpot’s most impactful options started as light-weight experiments or informal conversations inside the neighborhood.
Wanting Forward
What makes Penpot particular will not be solely its know-how however the way in which it’s constructed, with transparency, care, and a real perception in shared possession. Because the staff steps into the world of AI, that philosophy turns into much more vital. As an alternative of releasing a refined assistant, Penpot is creating house for exploration and collective invention.
The MCP Server is simply the start. The way forward for design tooling is not going to be outlined by a single firm or mannequin. Will probably be formed by the individuals who use these instruments every single day. Penpot is selecting to construct that future overtly, and in case you are interested in the place AI and design are headed, this is a perfect second to take part.
