# Dembrandt (npm · dembrandt) Design tokens from any live site: colours, type, spacing, components, drift against a baseline. - Trust score: 84/100 (high trust) - Change this week: +3 - Registry status: active - Liveness: live - Owner verified: no - Last scored: 2026-11-05 ## Components - npm · `dembrandt`: 84/100 (this document), [markdown](https://verifymcp.io/servers/dembrandt-dembrandt/dembrandt.md), [page](https://verifymcp.io/servers/dembrandt-dembrandt/dembrandt) ## Channel facts - Registry: `npm` - Package: `dembrandt` - Version: `0.39.2` - Transport: `stdio` ## Trust breakdown How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog) Scored 2026-11-05. - **Supply Chain Security**: 81/100 - No malware found by supply-chain analysis. - CVE check failed: a known high-severity CVE affects @modelcontextprotocol/sdk 1.30.0, a direct dependency. A fixed version is available. - No install/post-install scripts declared. - 42 of 127 dependencies flagged as unhealthy. - **Provenance & Transparency**: 100/100 - Source repository is publicly reachable at the declared URL. - Cryptographically verified build provenance (signed, bound to dembrandt/dembrandt). - Clear OSI-approved license (MIT). - Actively maintained (last published 24 days ago). - Publishes a security disclosure policy (SECURITY.md). - **Schema Quality & AI Usability**: 63/100 - AI-judged instruction clarity (excellent). - Context-footprint check failed: tool/resource definitions use about 5714 tokens (~272/item across 21 items; 21 tools + 0 resources), over budget; trim descriptions and params. - Usage-examples check failed: none of the tools include examples. - **Stability & Change Management**: 73/100 - Stability observed for 22 of 30 days with no destabilising changes; credit accrues until the full window elapses. - **Tool Coverage**: 100/100 - 100% of tools have a non-trivial description (not blank, and not just the tool's name). - 100% of tool parameters carry a description. - **Tool Safety**: 100/100 - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured. - We read all 21 captured tool definition(s), and no name or description among them implies an irreversible operation. - An AI judge read all 21 captured unit(s) of tool text and found none that tries to manipulate the model reading it. - **Capabilities**: 100/100 - Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28. ## Install ### How do I install the Dembrandt MCP server? Dembrandt runs locally as an npm package, launched with npx -y dembrandt mcp. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation. ### Claude ```bash claude mcp add dembrandt-dembrandt -- npx -y dembrandt mcp ``` ### Cursor ```json { "mcpServers": { "dembrandt-dembrandt": { "command": "npx", "args": [ "-y", "dembrandt", "mcp" ] } } } ``` ### VS Code ```json { "servers": { "dembrandt-dembrandt": { "command": "npx", "args": [ "-y", "dembrandt", "mcp" ] } } } ``` ### Codex ```bash codex mcp add dembrandt-dembrandt -- npx -y dembrandt mcp ``` ### opencode ```json { "$schema": "https://opencode.ai/config.json", "mcp": { "dembrandt-dembrandt": { "type": "local", "command": [ "npx", "-y", "dembrandt", "mcp" ], "enabled": true } } } ``` ### OpenClaw ```bash openclaw mcp add dembrandt-dembrandt --command npx --arg -y --arg dembrandt --arg mcp ``` ### Hermes ```yaml mcp_servers: dembrandt-dembrandt: command: "npx" args: ["-y", "dembrandt", "mcp"] ``` ### Netclaw ```json { "McpServers": { "dembrandt-dembrandt": { "Transport": "stdio", "Command": "npx", "Arguments": [ "-y", "dembrandt", "mcp" ] } } } ``` ### Vellum ```bash assistant mcp add dembrandt-dembrandt -t stdio -c npx -a -y dembrandt mcp ``` ### Other ```json { "mcpServers": { "dembrandt-dembrandt": { "command": "npx", "args": [ "-y", "dembrandt", "mcp" ] } } } ``` ## Changelog Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims. ### 2026-11-05 (score 84, +1) No change was recorded against any check on this day. Stability & Change Management went from 70 to 73. That category is still filling its 30-day observation window: 21 days of observed history at the previous scan, 22 at this one. The score rises as the window fills, whether or not the server changes. ### 2026-11-03 (score 83, +1) No change was recorded against any check on this day. Stability & Change Management went from 63 to 67. That category is still filling its 30-day observation window: 19 days of observed history at the previous scan, 20 at this one. The score rises as the window fills, whether or not the server changes. ### 2026-10-31 (score 82, +1) No change was recorded against any check on this day. Stability & Change Management went from 53 to 57. That category is still filling its 30-day observation window: 16 days of observed history at the previous scan, 17 at this one. The score rises as the window fills, whether or not the server changes. ### 2026-10-29 (score 81, +1) No change was recorded against any check on this day. Stability & Change Management went from 47 to 50. That category is still filling its 30-day observation window: 14 days of observed history at the previous scan, 15 at this one. The score rises as the window fills, whether or not the server changes. ### 2026-10-28 (score 80, 0) - [security regression] GHSA-f897-62hf-6gxq affects this package: high ### 2026-10-27 (score 80, +1) No change was recorded against any check on this day. Stability & Change Management went from 40 to 43. That category is still filling its 30-day observation window: 12 days of observed history at the previous scan, 13 at this one. The score rises as the window fills, whether or not the server changes. ### 2026-10-25 (score 79, +1) No change was recorded against any check on this day. Stability & Change Management went from 33 to 37. That category is still filling its 30-day observation window: 10 days of observed history at the previous scan, 11 at this one. The score rises as the window fills, whether or not the server changes. ### 2026-10-24 (score 78, 0) - [security regression] CVE-2023-54404 affects this package: high ## MCP tools (21) ### `get_design_tokens` (~575 tokens) Extract the full design system from a live website. Launches a real browser, navigates to the site, and returns production-ready design tokens: color palette (hex, RGB, LCH, OKLCH) with semantic roles and CSS custom properties, typography scale (families, fallbacks, sizes, weights, line heights, letter spacing by context), spacing system with grid detection, border radii, border patterns, box shadows for elevation, component styles (buttons with hover/focus states, inputs, links, badges), responsive breakpoints, logo and favicons, site name, the frameworks and libraries that build the UI, and icon systems. Set pages > 1 to crawl and merge several pages, which yields a markedly stronger token set than a single page. Returns a job_id by default: poll it with get_job_status, and pass the same job_id to compute_drift, get_findings, export_dtcg, generate_design_md or render_report instead of resending the extraction. Input parameters: - `basicAuth` (string): HTTP Basic credentials as "user:password" - `cookie` (string): Cookie string for authenticated pages, e.g. "session=abc; token=xyz" - `cookieFile` (string): Path to a cookie export: JSON array, Playwright storageState (cookies + localStorage), or Netscape cookies.txt - `darkMode` (boolean): Extract the dark theme: the page is rendered with prefers-color-scheme: dark - `header`: Extra HTTP header, e.g. "Authorization: Bearer eyJ...", or a list of them - `mobile` (boolean): Extract from a mobile viewport instead of desktop - `noSandbox` (boolean): Disable the browser sandbox, required inside Docker and most CI containers - `pages` (integer): Extract up to N pages and merge them into one token set. Pages are discovered from DOM links, or from sitemap.xml when sitemap is true. Merged tokens are markedly stronger than a single page. - `paths` (array): Explicit extra paths on the same domain to extract and merge, e.g. ["/pricing", "/docs"]. Overrides page discovery. - `sitemap` (boolean): Discover the extra pages from sitemap.xml instead of DOM links. Alone it takes up to 20 pages; set pages to cap it - `slow` (boolean): 3x timeouts for heavy SPAs - `sync` (boolean): Wait for the result directly instead of returning a job_id. Blocks 15-40s for one page, and proportionally longer for a multi-page crawl. - `url` (string, required): Website URL (e.g. example.com) - `userAgent` (string): Custom user agent string - `wcag` (boolean): Include WCAG contrast analysis between palette colors ### `get_color_palette` (~517 tokens) Extract brand colors from a live website. Returns semantic colors (primary, secondary, accent, plus background and text promoted from the page surface and body text), full palette ranked by usage frequency and confidence (high/medium/low), CSS custom properties with their design-system names, and hover/focus state colors discovered by simulating real user interactions. Each color in hex, RGB, LCH, and OKLCH. Set pages > 1 to crawl and merge several pages, which yields a markedly stronger token set than a single page. Returns a job_id by default: poll it with get_job_status, and pass the same job_id to compute_drift, get_findings, export_dtcg, generate_design_md or render_report instead of resending the extraction. Input parameters: - `basicAuth` (string): HTTP Basic credentials as "user:password" - `cookie` (string): Cookie string for authenticated pages, e.g. "session=abc; token=xyz" - `cookieFile` (string): Path to a cookie export: JSON array, Playwright storageState (cookies + localStorage), or Netscape cookies.txt - `darkMode` (boolean): Also extract dark mode palette - `header`: Extra HTTP header, e.g. "Authorization: Bearer eyJ...", or a list of them - `mobile` (boolean): Extract from a mobile viewport instead of desktop - `noSandbox` (boolean): Disable the browser sandbox, required inside Docker and most CI containers - `pages` (integer): Extract up to N pages and merge them into one token set. Pages are discovered from DOM links, or from sitemap.xml when sitemap is true. Merged tokens are markedly stronger than a single page. - `paths` (array): Explicit extra paths on the same domain to extract and merge, e.g. ["/pricing", "/docs"]. Overrides page discovery. - `sitemap` (boolean): Discover the extra pages from sitemap.xml instead of DOM links. Alone it takes up to 20 pages; set pages to cap it - `slow` (boolean): 3x timeouts for heavy SPAs - `sync` (boolean): Wait for the result directly instead of returning a job_id. Blocks 15-40s for one page, and proportionally longer for a multi-page crawl. - `url` (string, required): Website URL (e.g. example.com) - `userAgent` (string): Custom user agent string - `wcag` (boolean): Include WCAG contrast analysis between palette colors ### `get_typography` (~518 tokens) Extract typography from a live website. Returns every font family with its fallback stack, the complete type scale grouped by context (heading, body, text, button, link, caption) with pixel and rem sizes, weights, line heights, letter spacing, and text transforms. The body context marks the dominant reading-text font; text marks other body-eligible copy. Also reports font sources: Google Fonts URLs, Adobe Fonts usage, and variable font detection. Set pages > 1 to crawl and merge several pages, which yields a markedly stronger token set than a single page. Returns a job_id by default: poll it with get_job_status, and pass the same job_id to compute_drift, get_findings, export_dtcg, generate_design_md or render_report instead of resending the extraction. Input parameters: - `basicAuth` (string): HTTP Basic credentials as "user:password" - `cookie` (string): Cookie string for authenticated pages, e.g. "session=abc; token=xyz" - `cookieFile` (string): Path to a cookie export: JSON array, Playwright storageState (cookies + localStorage), or Netscape cookies.txt - `darkMode` (boolean): Extract the dark theme: the page is rendered with prefers-color-scheme: dark - `header`: Extra HTTP header, e.g. "Authorization: Bearer eyJ...", or a list of them - `mobile` (boolean): Extract from a mobile viewport instead of desktop - `noSandbox` (boolean): Disable the browser sandbox, required inside Docker and most CI containers - `pages` (integer): Extract up to N pages and merge them into one token set. Pages are discovered from DOM links, or from sitemap.xml when sitemap is true. Merged tokens are markedly stronger than a single page. - `paths` (array): Explicit extra paths on the same domain to extract and merge, e.g. ["/pricing", "/docs"]. Overrides page discovery. - `sitemap` (boolean): Discover the extra pages from sitemap.xml instead of DOM links. Alone it takes up to 20 pages; set pages to cap it - `slow` (boolean): 3x timeouts for heavy SPAs - `sync` (boolean): Wait for the result directly instead of returning a job_id. Blocks 15-40s for one page, and proportionally longer for a multi-page crawl. - `url` (string, required): Website URL (e.g. example.com) - `userAgent` (string): Custom user agent string ### `get_component_styles` (~498 tokens) Extract UI component styles from a live website. Returns button variants with default, hover, active, and focus states (background, text color, padding, border radius, border, shadow, outline, opacity), input field styles (border, focus ring, padding, placeholder), link styles (color, text decoration, hover changes), and badge/tag styles. Set pages > 1 to crawl and merge several pages, which yields a markedly stronger token set than a single page. Returns a job_id by default: poll it with get_job_status, and pass the same job_id to compute_drift, get_findings, export_dtcg, generate_design_md or render_report instead of resending the extraction. Input parameters: - `basicAuth` (string): HTTP Basic credentials as "user:password" - `cookie` (string): Cookie string for authenticated pages, e.g. "session=abc; token=xyz" - `cookieFile` (string): Path to a cookie export: JSON array, Playwright storageState (cookies + localStorage), or Netscape cookies.txt - `darkMode` (boolean): Extract the dark theme: the page is rendered with prefers-color-scheme: dark - `header`: Extra HTTP header, e.g. "Authorization: Bearer eyJ...", or a list of them - `mobile` (boolean): Extract from a mobile viewport instead of desktop - `noSandbox` (boolean): Disable the browser sandbox, required inside Docker and most CI containers - `pages` (integer): Extract up to N pages and merge them into one token set. Pages are discovered from DOM links, or from sitemap.xml when sitemap is true. Merged tokens are markedly stronger than a single page. - `paths` (array): Explicit extra paths on the same domain to extract and merge, e.g. ["/pricing", "/docs"]. Overrides page discovery. - `sitemap` (boolean): Discover the extra pages from sitemap.xml instead of DOM links. Alone it takes up to 20 pages; set pages to cap it - `slow` (boolean): 3x timeouts for heavy SPAs - `sync` (boolean): Wait for the result directly instead of returning a job_id. Blocks 15-40s for one page, and proportionally longer for a multi-page crawl. - `url` (string, required): Website URL (e.g. example.com) - `userAgent` (string): Custom user agent string ### `get_surfaces` (~475 tokens) Extract surface treatment tokens from a live website: border radii with element context (which radii are used on buttons vs cards vs inputs vs modals), border patterns (width + style + color combinations), and box shadow elevation levels. Set pages > 1 to crawl and merge several pages, which yields a markedly stronger token set than a single page. Returns a job_id by default: poll it with get_job_status, and pass the same job_id to compute_drift, get_findings, export_dtcg, generate_design_md or render_report instead of resending the extraction. Input parameters: - `basicAuth` (string): HTTP Basic credentials as "user:password" - `cookie` (string): Cookie string for authenticated pages, e.g. "session=abc; token=xyz" - `cookieFile` (string): Path to a cookie export: JSON array, Playwright storageState (cookies + localStorage), or Netscape cookies.txt - `darkMode` (boolean): Extract the dark theme: the page is rendered with prefers-color-scheme: dark - `header`: Extra HTTP header, e.g. "Authorization: Bearer eyJ...", or a list of them - `mobile` (boolean): Extract from a mobile viewport instead of desktop - `noSandbox` (boolean): Disable the browser sandbox, required inside Docker and most CI containers - `pages` (integer): Extract up to N pages and merge them into one token set. Pages are discovered from DOM links, or from sitemap.xml when sitemap is true. Merged tokens are markedly stronger than a single page. - `paths` (array): Explicit extra paths on the same domain to extract and merge, e.g. ["/pricing", "/docs"]. Overrides page discovery. - `sitemap` (boolean): Discover the extra pages from sitemap.xml instead of DOM links. Alone it takes up to 20 pages; set pages to cap it - `slow` (boolean): 3x timeouts for heavy SPAs - `sync` (boolean): Wait for the result directly instead of returning a job_id. Blocks 15-40s for one page, and proportionally longer for a multi-page crawl. - `url` (string, required): Website URL (e.g. example.com) - `userAgent` (string): Custom user agent string ### `get_spacing` (~466 tokens) Extract the spacing system from a live website: common margin and padding values sorted by frequency, pixel and rem values, and grid system detection (4px, 8px, or custom scale). Set pages > 1 to crawl and merge several pages, which yields a markedly stronger token set than a single page. Returns a job_id by default: poll it with get_job_status, and pass the same job_id to compute_drift, get_findings, export_dtcg, generate_design_md or render_report instead of resending the extraction. Input parameters: - `basicAuth` (string): HTTP Basic credentials as "user:password" - `cookie` (string): Cookie string for authenticated pages, e.g. "session=abc; token=xyz" - `cookieFile` (string): Path to a cookie export: JSON array, Playwright storageState (cookies + localStorage), or Netscape cookies.txt - `darkMode` (boolean): Extract the dark theme: the page is rendered with prefers-color-scheme: dark - `header`: Extra HTTP header, e.g. "Authorization: Bearer eyJ...", or a list of them - `mobile` (boolean): Extract from a mobile viewport instead of desktop - `noSandbox` (boolean): Disable the browser sandbox, required inside Docker and most CI containers - `pages` (integer): Extract up to N pages and merge them into one token set. Pages are discovered from DOM links, or from sitemap.xml when sitemap is true. Merged tokens are markedly stronger than a single page. - `paths` (array): Explicit extra paths on the same domain to extract and merge, e.g. ["/pricing", "/docs"]. Overrides page discovery. - `sitemap` (boolean): Discover the extra pages from sitemap.xml instead of DOM links. Alone it takes up to 20 pages; set pages to cap it - `slow` (boolean): 3x timeouts for heavy SPAs - `sync` (boolean): Wait for the result directly instead of returning a job_id. Blocks 15-40s for one page, and proportionally longer for a multi-page crawl. - `url` (string, required): Website URL (e.g. example.com) - `userAgent` (string): Custom user agent string ### `get_motion` (~464 tokens) Extract the motion system from a live website: transition and animation durations with their usage counts, easing curves, the durations and easings used per component context (button, link, nav, card, modal), named keyframe animations, and the hover patterns discovered by simulating real interaction. Also returns gradients, which travel with motion as the decorative layer. Set pages > 1 to crawl and merge several pages, which yields a markedly stronger token set than a single page. Returns a job_id by default: poll it with get_job_status. Input parameters: - `basicAuth` (string): HTTP Basic credentials as "user:password" - `cookie` (string): Cookie string for authenticated pages, e.g. "session=abc; token=xyz" - `cookieFile` (string): Path to a cookie export: JSON array, Playwright storageState (cookies + localStorage), or Netscape cookies.txt - `darkMode` (boolean): Extract the dark theme: the page is rendered with prefers-color-scheme: dark - `header`: Extra HTTP header, e.g. "Authorization: Bearer eyJ...", or a list of them - `mobile` (boolean): Extract from a mobile viewport instead of desktop - `noSandbox` (boolean): Disable the browser sandbox, required inside Docker and most CI containers - `pages` (integer): Extract up to N pages and merge them into one token set. Pages are discovered from DOM links, or from sitemap.xml when sitemap is true. Merged tokens are markedly stronger than a single page. - `paths` (array): Explicit extra paths on the same domain to extract and merge, e.g. ["/pricing", "/docs"]. Overrides page discovery. - `sitemap` (boolean): Discover the extra pages from sitemap.xml instead of DOM links. Alone it takes up to 20 pages; set pages to cap it - `slow` (boolean): 3x timeouts for heavy SPAs - `sync` (boolean): Wait for the result directly instead of returning a job_id. Blocks 15-40s for one page, and proportionally longer for a multi-page crawl. - `url` (string, required): Website URL (e.g. example.com) - `userAgent` (string): Custom user agent string ### `get_brand_identity` (~500 tokens) Extract brand identity from a live website: site name, logo (source, dimensions, safe zone), all favicon variants (icon, apple-touch-icon, og:image, twitter:image with sizes and URLs), what builds the UI with versions where recoverable (JS and meta frameworks, CSS frameworks, component libraries, site builders), icon systems, and responsive breakpoints. Set pages > 1 to crawl and merge several pages, which yields a markedly stronger token set than a single page. Returns a job_id by default: poll it with get_job_status, and pass the same job_id to compute_drift, get_findings, export_dtcg, generate_design_md or render_report instead of resending the extraction. Input parameters: - `basicAuth` (string): HTTP Basic credentials as "user:password" - `cookie` (string): Cookie string for authenticated pages, e.g. "session=abc; token=xyz" - `cookieFile` (string): Path to a cookie export: JSON array, Playwright storageState (cookies + localStorage), or Netscape cookies.txt - `darkMode` (boolean): Extract the dark theme: the page is rendered with prefers-color-scheme: dark - `header`: Extra HTTP header, e.g. "Authorization: Bearer eyJ...", or a list of them - `mobile` (boolean): Extract from a mobile viewport instead of desktop - `noSandbox` (boolean): Disable the browser sandbox, required inside Docker and most CI containers - `pages` (integer): Extract up to N pages and merge them into one token set. Pages are discovered from DOM links, or from sitemap.xml when sitemap is true. Merged tokens are markedly stronger than a single page. - `paths` (array): Explicit extra paths on the same domain to extract and merge, e.g. ["/pricing", "/docs"]. Overrides page discovery. - `sitemap` (boolean): Discover the extra pages from sitemap.xml instead of DOM links. Alone it takes up to 20 pages; set pages to cap it - `slow` (boolean): 3x timeouts for heavy SPAs - `sync` (boolean): Wait for the result directly instead of returning a job_id. Blocks 15-40s for one page, and proportionally longer for a multi-page crawl. - `url` (string, required): Website URL (e.g. example.com) - `userAgent` (string): Custom user agent string ### `compute_drift` (~255 tokens) Compare two dembrandt extractions and return a design-drift report: a 0-100 score (0 = identical), a stable/drift verdict, per-category scores, and the list of changed/added/removed tokens (colors, typography, spacing, radius, shadows). Pure and synchronous, no browser. Takes either an inline extraction or the job_id of a completed one. Use it to check whether generated or updated UI has drifted from a brand baseline. Input parameters: - `baseline` (object): A dembrandt extraction object. Omit it and pass job_id instead to read a completed extraction straight out of the job queue, which avoids sending the whole extraction back through the model. - `baselineJobId` (string): job_id of a completed extraction to read instead of passing result inline - `candidate` (object): A dembrandt extraction object. Omit it and pass job_id instead to read a completed extraction straight out of the job queue, which avoids sending the whole extraction back through the model. - `candidateJobId` (string): job_id of a completed extraction to read instead of passing result inline - `failThreshold` (number): Score above this yields a 'drift' verdict (default 10) ### `render_report` (~153 tokens) Render a self-contained HTML report (inline CSS, no external resources) from a dembrandt extraction, optionally including a drift diff. Takes either an inline extraction or the job_id of a completed one. Returns the HTML as text: write it to a .html file to open offline or attach as a CI artifact. Input parameters: - `drift`: A drift report from compute_drift, to render the diff banner - `job_id` (string): job_id of a completed extraction to read instead of passing result inline - `result` (object): A dembrandt extraction object. Omit it and pass job_id instead to read a completed extraction straight out of the job queue, which avoids sending the whole extraction back through the model. ### `get_findings` (~172 tokens) Lint a dembrandt extraction for design-system quality issues: WCAG contrast failures, inconsistency (near-duplicate colors, off-scale spacing values, radius sprawl), and duplication. Returns findings with severity (error/warn), category, and a human-readable message, plus summary counts. Pure and synchronous, no browser. Takes either an inline extraction or the job_id of a completed one. Complements compute_drift: drift asks 'did it change', findings asks 'is it good'. Input parameters: - `job_id` (string): job_id of a completed extraction to read instead of passing result inline - `result` (object): A dembrandt extraction object. Omit it and pass job_id instead to read a completed extraction straight out of the job queue, which avoids sending the whole extraction back through the model. ### `export_dtcg` (~162 tokens) Convert a dembrandt extraction to W3C Design Tokens (DTCG) format: color, typography, spacing, radius, border, and shadow tokens with $type/$value structure and dembrandt provenance under $extensions. Pure and synchronous, no browser. Takes either an inline extraction or the job_id of a completed one. Use it to hand tokens to Style Dictionary, Figma token plugins, or any DTCG-compatible pipeline. Input parameters: - `job_id` (string): job_id of a completed extraction to read instead of passing result inline - `result` (object): A dembrandt extraction object. Omit it and pass job_id instead to read a completed extraction straight out of the job queue, which avoids sending the whole extraction back through the model. ### `generate_design_md` (~146 tokens) Render a DESIGN.md brand guide (markdown) from a dembrandt extraction: colors, typography, spacing, surfaces, and components as a human-readable design reference. Pure and synchronous, no browser. Takes either an inline extraction or the job_id of a completed one. Write the output to DESIGN.md in a project so agents and developers build UI against the extracted brand. Input parameters: - `job_id` (string): job_id of a completed extraction to read instead of passing result inline - `result` (object): A dembrandt extraction object. Omit it and pass job_id instead to read a completed extraction straight out of the job queue, which avoids sending the whole extraction back through the model. ### `export_tailwind` (~151 tokens) Render a Tailwind v4 @theme CSS block from a dembrandt extraction: colors, typography, spacing, radii and shadows as custom properties, observed values only, with nothing invented. Pure and synchronous, no browser. Takes either an inline extraction or the job_id of a completed one. Write the output to a project's CSS entry point so Tailwind utilities resolve to the measured brand. Input parameters: - `job_id` (string): job_id of a completed extraction to read instead of passing result inline - `result` (object): A dembrandt extraction object. Omit it and pass job_id instead to read a completed extraction straight out of the job queue, which avoids sending the whole extraction back through the model. ### `export_shadcn` (~169 tokens) Render a shadcn/ui theme from a dembrandt extraction: the :root block and the @theme inline mapping Tailwind v4 needs. A slot is written only where the page supplied a value, and the rest are named in the file header and left at shadcn's own defaults, so no slot is filled with an invented value that reads as measured. Pure and synchronous, no browser. Takes either an inline extraction or the job_id of a completed one. Input parameters: - `job_id` (string): job_id of a completed extraction to read instead of passing result inline - `result` (object): A dembrandt extraction object. Omit it and pass job_id instead to read a completed extraction straight out of the job queue, which avoids sending the whole extraction back through the model. ### `validate_dtcg` (~120 tokens) Validate a W3C Design Tokens (DTCG) document against the 2025.10 spec: token types, colour objects and their component ranges, dimensions, references, and property-level $ref pointers. Returns valid plus the list of errors with the path each one sits at. Pure and synchronous, no browser. Use it after writing or editing a token file, including one this server produced, so a hand edit cannot quietly break the document. Input parameters: - `tokens` (object, required): The DTCG token document to validate, as an object ### `check_contrast` (~133 tokens) Grade colour pairs against WCAG 2.1 contrast, at the threshold the text size earns: 18pt, or 14pt bold, is large scale and needs 3:1 where body text needs 4.5:1. Takes pairs you name, so it grades colours you are about to ship rather than only colours already on a page. Returns the ratio, the required ratio, and the AA and AAA verdicts per pair. Pure and synchronous, no browser. For pairs already rendered on a site, extract with wcag instead. Input parameters: - `pairs` (array, required): The colour pairs to grade ### `check_robots` (~92 tokens) Ask whether robots.txt allows extracting a URL, before spending a browser run on it. Returns the verdict, the rule that decided it, and the robots.txt status. A 404 or 410 means no robots.txt and everything is allowed; any other unreadable response is treated as a refusal. Cheap and synchronous: one HTTP request, no browser. Input parameters: - `url` (string, required): The URL you intend to extract ### `get_job_status` (~68 tokens) Poll for the result of an async extraction job. Returns status (queued/running/completed/failed/cancelled) and the full result once completed. Call this after any extraction tool that returned a job_id. Input parameters: - `job_id` (string, required): The job_id returned by an extraction tool ### `list_jobs` (~43 tokens) List all extraction jobs from this session with their status (queued/running/completed/failed/cancelled), URL, and timestamps. Completed jobs are kept for one hour. ### `cancel_job` (~37 tokens) Cancel a queued extraction job. Has no effect on jobs that are already running. Input parameters: - `job_id` (string, required): The job_id to cancel ## Diagnostics Captured diagnostic sections: Provenance, Vulnerabilities, Dependencies. The full working is on the page: https://verifymcp.io/servers/dembrandt-dembrandt/dembrandt#diagnostics ## Score history - 2026-11-05: 84 - 2026-11-04: 83 - 2026-11-03: 83 - 2026-11-02: 82 - 2026-11-01: 82 - 2026-10-31: 82 - 2026-10-30: 81 - 2026-10-29: 81 - 2026-10-28: 80 - 2026-10-27: 80 - 2026-10-26: 79 - 2026-10-25: 79 - 2026-10-24: 78 - 2026-10-23: 78 - 2026-10-22: 77 - 2026-10-21: 74 - 2026-10-20: 74 - 2026-10-19: 74 - 2026-10-18: 74 - 2026-10-17: 74 - 2026-10-16: 74 - 2026-10-15: 74 - 2026-10-14: 74 - 2026-10-13: 44 - 2026-10-12: 44 ## Common questions ### What is the Dembrandt MCP server? Dembrandt is an MCP server listed in the public MCP registry as io.github.dembrandt/dembrandt. Design tokens from any live site: colours, type, spacing, components, drift against a baseline. This page covers its npm package (dembrandt). ### Is the Dembrandt MCP server safe to use? Dembrandt scores 84 out of 100 on VerifyMCP. We recorded 7 known advisories against it as of 5 November 2026. It declares no install or post-install scripts. Its build provenance is signed and verified. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm. ### What tools does the Dembrandt MCP server expose? Dembrandt exposes 21 tools: get_design_tokens, get_color_palette, get_typography, get_component_styles, get_surfaces, and 16 more. Their descriptions and schemas cost roughly 5,714 tokens of context every time the server is loaded. ### Is the Dembrandt MCP server still maintained? Dembrandt is still listed as active in the MCP registry. We last reached this channel on 5 November 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced. ### What licence is the Dembrandt MCP server under? Dembrandt declares the MIT licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls. ### What did VerifyMCP test on the Dembrandt MCP server? VerifyMCP ran 16 checks on Dembrandt, covering Malware scan, Known CVEs, Install scripts, Dependency health and 12 more. Endpoint Security and Transport & Reachability do not apply to an npm package, so they were not tested. The score does not measure everything, and the scoring documentation lists what it leaves out. ## Links - npm package: https://www.npmjs.com/package/dembrandt - Socket report: https://socket.dev/npm/package/dembrandt - Repository: https://github.com/dembrandt/dembrandt - Website: https://dembrandt.com/ - Changelog RSS feed: https://verifymcp.io/servers/dembrandt-dembrandt/dembrandt.xml - Changelog JSON feed: https://verifymcp.io/servers/dembrandt-dembrandt/dembrandt/changelog.json - Trust report (JSON): https://verifymcp.io/servers/dembrandt-dembrandt/dembrandt.json - HTML version of this page: https://verifymcp.io/servers/dembrandt-dembrandt/dembrandt