Suggest a Feature

I use Claude on two separate Claude accounts, and both need the Figma MCP connector pointed at the same Figma account. This does not work. As soon as I authorize the Figma connector on the second Claude account, the first one is invalidated and its Figma tools stop working (“The user’s connection to this connector was invalidated”). Re-authorizing on the first account then kicks off the second. The same Figma login can be authorized in Claude Code’s own Figma plugin at the same time without conflict, so the limit seems to be one active authorization per Figma account for the Claude connector client, rather than a general one-session rule. Please allow one Figma account to hold more than one active authorization for the same MCP client, or document the limit and the intended workaround. Reproduced today on the current remote server at mcp.figma.com with two Claude accounts, one Figma account.

Hi Figma team,Our team uses Storybook to implement and document our components and pages in code. Doing this manually takes developers a lot of time: they have to rebuild every Component and Variant in code and write a story for each State.So we built our own agent that converts our Figma designs into Storybook. The results were great: it dramatically sped up our workflow, and the generated code ended up much closer to the original design.To give you a real picture, here’s what we built in our project in just one week:UI Kit: 27 component sets, 2,288 variants, 4 Sidebar components and 1 Chart instance Icons: 1,254 icon sets with 7,520 style variants, plus 259 flags and 3 standalone assets Style guide: 89 color styles and 42 typography styles Design variables: 157 across Primitives (16), Tokens (34), Colors (89) and Typography (18)For a senior developer, implementing all of this manually would have taken around 2–3 months.That experience gave me this idea: it would be amazing if an agen

Your improvements to auto layout are incredible. I can’t tell you how much I love seeing the Figma team tackle these core features. There is a “final boss” of the Grid auto-layout that I’d like to suggest potentially tackling. It looks like you’re heading in this direction, so please carry on if you’ve already started :-)“repeat(auto-fit, minmax())” Responsive grid columns, equivalent to repeat(auto-fit, minmax()).Expose a minimum card/column width in the GUI, with the grid automatically reducing its column count and adding rows as the container narrows. This would make a powerful—but challenging—CSS pattern much more accessible to designers and some developers. 

There’s no visible scrollbar in Figma prototyping mode. Yesterday we’re having user testings and once again this same topic popped out. Some of the users didn’t thought that they could scroll prototype because scrollbar wasn’t visible. One of the user was using laptop’s touchpad and he wasn’t able to scroll with only using touchpad. Guiding user how to use prototype during test can harm testing situations. Atleast it’s sure that if user thought that he can’t scroll it leads to biased feedback about product. If there’s some reason not to have scrollbar it could be atleast optional feature that designer could turn visible scrollbar on from options. It’s not only who has came across with this weakness in Figma. Here’s old Spectrum thread about same topic but Figma hasn’t done anything for this yet and I hope this get more attention here.

Often I’m adding autolayout around a frame/component/element, and the default vertical and horizontal padding is 10px. As well, the gap applied is 10px. It would be amazing to set the default padding and gap to auto-layout elements. This would help with keeping your files clean, but also, to ensure that redundant CSS properties are not added to dev mode output. Clearing and resetting these properties takes time and often the gap is not cleared due to the property not having an affect on single children. Thanks!

The current system for managing team members/seats and understanding file access and permissions is wildly fractured and convoluted. Currently, there is the Admin Dashboard, where I can manage seat type, but not see file access; the Team Members panel, where I can see what type of access people have (Member or Limited Access) but am NOT able to change access; and individual People/Profile pages where I can see file contributions, but this requires searching for the person using the file browser, and cannot be accessed through the aforementioned dashboard/panel (? why).Right now, finding details about seats, file access, and file contributions is three step process involving all of the above. A singular admin dashboard where you can see team members, access, seat type, and have a comprehensive view of file access/contributions for each user would make life so much easier. I wouldn’t be surprised to find out this is available and paywalled behind a more expensive license, but to me seems

Figma is already a great tool for designers, but one thing it is missing is CMYK support for print design. Many of us work on both digital and print projects, and having to switch to other software just for CMYK is not very convenient. If Figma adds CMYK support, we could do almost everything in one place. I think this could be a game changer for Figma and give it a big advantage in the design market. Hope the Figma team considers this feature!

I am able to create a table using the grid auto-layout option. but I’m not able to reorder the columns or rows. It would be a great feature.

I just want to reinforce what Luis posted last year, since I agree 100% with him. Original thread:  It would be incredibly helpful if there were a way to include a rejection reason when declining a Figma seat upgrade request.I frequently receive upgrade notifications from across the organization, and in most cases, the requestors are unaware that there's an internal approval process they need to follow, such as submitting a request through our IT department rather than requesting directly via Figma.Lately, this has become quite challenging to manage. When I reject these requests, I often receive frustrated follow-up emails from users who don’t understand why their upgrade was denied. Being able to provide a brief explanation at the time of rejection would save time, reduce confusion, and help streamline communication.Thanks for considering this! 

Now that the glass effect has been released (which, by the way, is amazing), it's time to remember the 9-slice feature, which still has not been implemented natively. I'm talking about functionality similar to what exists in Unity. In this system, an image is marked with guides — without being sliced into separate parts — so that it can be stretched without distorting the corners. Unity, by the way, allows not only stretching the middle sections but also tiling them. Yes, it's possible to achieve this functionality now using plugins in Figma, but all of them work by slicing the image into parts, which makes updating these images complicated. It's often easier to create a new one than to edit an existing one.Since this functionality is very common in game development and mobile game design, it would be great to have compatibility with Unity for this feature. That way, the same image could be used both in Unity and Figma. This would greatly simplify the pipelines and effectiveness of gam

I’d like to organize the left-hand “Pages” panel by adding color codes to each page. (Think how Mac’s OSX Finder allows you to add colored circles to individual files.) Additionally, Pages should have the ability to be sorted into groups within a parent folder. (Think how you can make groups in Photoshop’s layers panel) This feature will make it easier for my team to organize our Experience Library file. Users need to be able to scan the Pages quickly. I’ve attached an example of what the color coding could look like.

Current music is repeating. I get boring after listening in 30 minutes and open some lo-fi hip hop stream. Really want to use FigJam music. Please, add more tracks or direct audio stream from SOME lo-fi music stream. Sincerely, Mikhail.

SummaryI'd like Mistral's MCP client to be added to the Figma MCP Catalog so that Mistral users can connect to the official remote Figma MCP server (https://mcp.figma.com/mcp).The problemMistral's Le Chat, now part of Mistral Vibe (Vibe Work), supports custom MCP connectors. These connect to remote servers over Streamable HTTP and support OAuth 2.1 with dynamic client registration. When I add https://mcp.figma.com/mcp as a custom connector, [describe what happened, e.g. the connection fails with "MCP connection requires additional information or is invalid" / HTTP 403 during registration].From the forum thread "Figma's approach breaks the core promise of MCP", I understand the remote server allowlists client_name during dynamic client registration, and that only clients in the MCP Catalog can currently authenticate. Mistral's client doesn't appear to be on that list.Why it matters- Mistral is a major AI provider with a large user base, particularly in Europe, where many teams choose it

This is a suggestion for a “quality of life improvement” new feature. Current behaviour:The variable selection within a collection and its extended collections are not synced.When one selects a variable in an collection / extended collection, the variable selection can be different for a collection and its extended collection.Steps to recreate current behaviour: Select collection P Select variable A within that collection Select extended collection (P extended) Select another variable B within the extended collection Switch back to collection P The selected variable in collection P is still AExpected behaviour:The expected behaviour is that the selected variable should be synced between a collection and its extended collections.Typical flow of a designer is to extend a collection and then modify specific variables within the extended collection. They switch back and forth between the collections to validate the previous variable value and the modified variable value. Steps to recreate

Figma motion does not currently have the ability to play and export video fills, so video objects are treated as static images. The ability to incorporate video objects into the timeline would be extremely helpful. Exposing the current frame of a video as a keyframeable property be a great solve. 

This would be handy if I want to mass-select and find hidden layers to show altogether. Unless this is a technical limitation, I am shocked that this is not integrated as an option from the get-go.

Hello Awesome community, It would be very helpful for Design System builder community support if we could create styles based on other styles, aka style variants to support use cases like these 3: Type Style A needs to be used with an underline to emphasize that the text is a hyperlink. There is no way to amend the style A without unlinking it at the moment. So either I create an extra style or I have elements in my design system that are not ruled by Type Styles; in both cases this creates extra work and potential for errors when updating the typography for specific projects based on the Design system file. Color Style B is used as a Primary accent color @100% opacity but also as a SurfaceTints and State Brushes with varying opacity values. This requires a design system file to include individual color styles for each case (aka Brushes based off of Primary) as parent containers currently set opacity for all content unless said opacity is par of a Color Style. This is extremely tim

Hi! Has the canvas panning keyboard shortcuts changed recently? I use Ctrl+Space+scroll wheel to pan and zoom the canvas but now that keyboard shortcut puts me into a Keyboard picking mode. Is this new? It really gets in the way of working.

Im creating a prototype and would like to link to live pages on our site without having the User in the test have to see the “Leaving Figma Warning” page. My concern is that it takes it breaks the illusion of an actual experience. Is there a way remove this middle page from the experience?

Currently, dashed strokes have ‘trimmed caps’. Is there a way to have all dashes with the same length? Example:

Currently, I have to manually scroll through the Recent tools list every time to find a tool I use frequently. This becomes time-consuming and inefficient. 

Description:When using Find and Replace in Figma, the search scope (Search in this file / Search in all pages) can be clearly visible and configurable before the search begins.Currently, if “Search in all pages” remains selected from a previous search, users may unintentionally perform a broader search than intended. Suggested Improvement: Display the search scope options before executing the search. Make the currently selected scope clearly visible. Allow users to switch between “Search in this file” and “Search in all pages” before execution. Consider defaulting to the current page/file scope when appropriate, rather than retaining a broader scope without sufficient visibility.   Rationale:This would improve user control, prevent unintended broad searches, and potentially reduce unnecessary processing time in large files. Expected Outcome:Users can make an informed scope selection before each search, improving predictability, efficiency, and overall usability. 

Hey I have a style guide and its currently in phase 2, my team has decided to simplify the text styles so I need to merge a lot of text styles together. Is there an efficient method of doing this? Thanks!

I'm excited for Bulk Create in Figma Buzz. It’s a lot more intuitive for our marketers to use, but at its current stage it’s actually more time consuming than our existing workflow, mainly because it doesn’t have the ability to bulk rename frames or the exported filenames. That means we still end up having to manually rename "Frame1.png" all the way to "Frame175.png".Here’s our use case: we need to create OpenGraph images for a website with over a hundred pages. We have two design templates: one with page title only, one with page title and a subhead.Currently, the marketing & web dev team maintain a spreadsheet of all the pages currently exist.We have the 2 designs set up as components and use a plugin called Scribe - Dynamic data in seconds to pull in the CSV and use a column we choose for naming the frames / files.I can see Bulk Create doing this as well, and even adding some common renaming & conditional naming options (i.e. case options, removing + replacing + appending ch