Query a Figma subtree by pattern—skeleton, buttons, text by font size, components, or semantic roles—instead of dumping the entire tree.
MIT
359,354 tools. Last updated 2026-08-01 10:23
"Plume" matching MCP tools:
- Runs heuristic accessibility checks on a PDS snapshot, identifying text contrast issues against resolved background color and button touch-target size violations.MIT
- Export Figma assets as SVG icons and PNG images via a paired plugin. Supports default, list, and surgical modes for targeted or broad exports.MIT
- Find Figma nodes by name and type across pages. Use queries like 'primary button' or 'TEXT layers named Title' to locate elements with details such as id, name, type, page, and size.MIT
- Compare two PDS snapshots of the same screen to detect semantic design changes. Returns structured deltas with role-labeled notes and a summary, avoiding raw JSON diffs.MIT
- Validate and compile a DSL document into Figma auto-layout frames with text, images, and icons. Supports create, sync, and dry-run modes with optional design brief intake.MIT
Matching MCP Servers
- AlicenseAqualityBmaintenanceAI-native design engineering platform that converts Figma or live websites into a semantic design graph for code generation, and generates Figma designs from prompts with a self-improving director loop.Last updated2819168MIT
- Alicense-qualityBmaintenancePlumb is an MCP server that provides coding agents with IDE-level intelligence (LSP, tree-sitter index) and safe, atomic file writes with transactional rollback and crash resilience.Last updated2MIT
- Extract a Figma screen or node as a normalized design spec with deduplicated tokens and CSS-shaped node tree, auto-layout resolved to flexbox, for code generation.MIT
- Self-critiques Figma designs against authored intent and design rubrics, scoring structure, design, and optional vision grade, then coaches iterative refinements to meet quality targets.MIT
- Check the current status of Plumb integration: view available data paths (Figma plugin or REST token), number of screens detected, compact-key legend for PDS responses, and token budget. No Figma access required.MIT
- Converts a design graph from Figma or a live site into pixel-faithful React JSX with inline styles, using a template-based deterministic emitter.MIT
- Creates a brand board from a brief: discovers and screenshots reference sites, synthesizes a color palette and type scale, and builds the result into a paired Figma file.MIT
- Extract current Figma selection data for code generation. Returns a compact semantic design graph without needing file keys or tokens.MIT
- Read a .fig file from disk to list all screens across every page. Use it to inspect or implement screens without opening Figma, requiring no plugin pairing or token.MIT
- Read a node from a .fig file by its sessionID:localID id (from plumb_fig_outline), returning type, name, size, fills, opacity, and text details.MIT
- Import a live webpage's visible DOM structure and semantics, mapping it onto a semantic design graph with real CSS values and detected roles. Use to audit site structure or track changes over time with multi-viewport support.MIT
- Scan live reference URLs to extract structural style digests (nav, hero, footer, card) before generating designs, ensuring new sections resemble reference layouts and typography.MIT
- Generate a full landing page in Figma from a one-line brief. Discovers reference sites, synthesizes a brand palette, and builds nav, hero, features, gallery, content split, CTA, and footer with derived copy.MIT
- Extract a deduplicated design-token table from a Figma node, including colors, typography, radii, and shadows, as $-prefixed refs for building with consistent tokens.MIT
- Generate a textual description of a Figma screen or node, breaking down layout regions and child elements, for use when rendered images are inaccessible.MIT
- List every Figma component definition and its instance usages to understand a design system before building, with an optional health report detecting unused components, duplicates, and variant outliers.MIT