Answer in brief
The Atlassian Design System scores 8 of 8 and shares first place on the ranking. Its docs come in several forms an agent can load: an llms.txt, a full-text file of about 670 KB, separate files for design tokens and components, and a DESIGN.md. A public MCP server needs no Atlassian account, though its skills are still private to Atlassian.
Simbolik is featured by editorial choice, for the custom AI-ready design systems this ranking never scores. Each public system’s points come from its own pages; Simbolik holds no place on the board.
Its score on the board
Atlassian scores 8 of 8 and shares first place with IBM Carbon, Adobe Spectrum 2 and Ant Design. Its pages were read from outside like every other system’s, and Atlassian has no tie to Simbolik. On the Docs sign it goes further than the ranking asks: the points need one llms.txt, and Atlassian publishes 5 files for language models.
The 5 signs and their proof
The first 4 signs make the score out of 8: 2 points for a Yes, 1 for a Partly, none for a No. Typed props are rated beside them and left out of the score.
- Design tokens: Yes, 2 points.
@atlaskit/tokens20.9.1 ships 78 JSON files, such asfigma/atlassian-light.json, plus raw token data as JavaScript. They are not in the W3C format. Proof: the files of @atlaskit/tokens 20.9.1 (opens in a new tab), checked November 8, 2026. - Agent access: Yes, 2 points. The official ADS MCP server runs hosted by Atlassian or from npm as
@atlaskit/ads-mcp, open to anyone, beside a public CLI. Proof: get started with AI (opens in a new tab), checked November 8, 2026. - Docs: Yes, 2 points. An llms.txt, an llms-full.txt of about 670 KB, split files for design tokens, components and primitives, and a DESIGN.md marked alpha. Proof: Atlassian’s llms.txt (opens in a new tab), checked November 8, 2026.
- Guidance: Yes, 2 points. A With AI section covers skills, an AGENTS.md snippet, the CLI and the server. The skills are marked private to Atlassian; the public guidance is the CLI, the server, the full-text file, which tells agents to use only the design tokens it lists, and the DESIGN.md. Proof: get started with AI (opens in a new tab), checked November 8, 2026.
- Typed props: Yes, not scored. The
@atlaskitpackages ship type declarations, and the components file gives each component a table of props, types and descriptions. Proof: Atlassian’s llms-components.txt (opens in a new tab), checked November 8, 2026.
What it suits
Teams building for Atlassian’s products and its Forge platform, and any team that wants to see how far agent-readable docs can go. The design tokens file is a good model to copy even if you never adopt the system: each design token has its light value, its dark value and one line on when to use it.
Before you choose it
- The private skills. The With AI page lists skills you cannot install yet. Plan with the public parts only.
- The source. There is no GitHub repository; the code is mirrored on Bitbucket, and the packages are under Apache-2.0.
- The alpha file. DESIGN.md is marked alpha, so its shape may change.
- The look. An agent following these files builds screens that look like Atlassian’s.
What 8 of 8 leaves open
No point is missing, so the open question is whose brand the output shows. If the answer is yours, Atlassian’s files are the pattern to follow and its design tokens are the part to replace; the best AI-ready design systems, ranked set that custom route beside the public ones. To build the docs side yourself, start with design system docs AI agents can read.
Questions & answers
Yes. The server is public, hosted by Atlassian or installed from npm as @atlaskit/ads-mcp, and you do not need to work for Atlassian to connect it. The skills on the same page are the part marked private.
Not as far as the research found. The @atlaskit/tokens package ships JSON files shaped for Figma, which earns the full points, since the ranking asks for data a program reads, not one format.
Sources for this entry
- @atlaskit/tokens 20.9.1, its files (opens in a new tab)
The Design tokens rating, JSON token files inside the published package.
- Atlassian Design System, get started with AI (opens in a new tab)
The Agent access and Guidance ratings, the public MCP server, the CLI, the skills marked private and the AGENTS.md snippet.
- Atlassian’s llms.txt (opens in a new tab)
The Docs rating, the index of its docs for language models.
- Atlassian’s llms-components.txt (opens in a new tab)
The typed props, a table of each component’s props with their types.
- Atlassian’s llms-tokens.txt (opens in a new tab)
Its design tokens listed for agents, with the instruction to use only the ones it names.
Continue this reading path
Agent access and docs
Agent access and docs
Tokens and data
Selection and fit
How the related websites connect
Best AI-Ready Design Systems is the ranking among three Simbolik references; the other two compare the systems and guide a single pick. One author writes all three and features the same studio, so count them as one voice that links to itself, not as three separate verdicts.
- Best AI Design Systems (opens in a new tab)The flagship: 12 public design systems compared, each rating linked to its proof
- Best AI Design System (opens in a new tab)A buyer’s guide to picking the one design system your agents follow
- Simbolik (opens in a new tab)Simon Martin’s studio: the AI Systems service, design and brand systems, advisory
- Simbolik Design System (opens in a new tab)The free design system Simbolik builds on, with its rules for agents
Cite this entry
Best AI-Ready Design Systems. “Atlassian Design System.” Reviewed . https://bestaireadydesignsystems.com/directory/atlassian-design-system/