Architecture & Design MCP servers

The MCP server connects your AI assistant straight to SVGator, the online SVG animation tool. From there, your assistant can work inside your projects to browse, inspect, edit, create, and export them for you, handling everything from a simple “edit colors according to this brand kit” to a complete “build me an animated logo reveal and export it as a CSS-only file.” - Endpoint: https://mcp.svgator.com/mcp - Protocol: Model Context Protocol over Streamable HTTP - Authentication: OAuth 2.1 (sign in with your SVGator account — see Getting started)

6 tools · create_project, edit_part

A lightweight MCP (Model Context Protocol) server for Blender. It offers a natural language interface with Blender’s Python API, improving access to documentation, and allowing users to explore and understand complex setups. For technical details and documentation about the architecture check the source code. > [!warning] Warning > Security Warning > The MCP server will execute LLM generated code in Blender without any guards in place to protect your data from removal or being sent to a remote location. To keep your data safe it is recommended to use a virtual machine, or a system without access to sensitive information.

5 tools · Data-block renaming, Geometry Nodes documentation

The Compeller MCP endpoint implements the Model Context Protocol as a thin JSON-RPC 2.0 wrapper over the existing v1 REST API. It is intended for agent integrators (Claude Desktop, Cursor, custom MCP clients, DigiRAMP) that speak MCP natively rather than raw HTTP. - Transport: Streamable HTTP (single JSON-RPC message per HTTP POST). - URL: POST https://compeller.ai/api/mcp - Protocol version: 2024-11-05 - Server name / version: compeller-mcp / see initialize result. - Tool contract: The tool list below is the public integration contract. Use tools/list for the runtime-advertised set on the deployed server. - Directory listings: Official MCP Registry · Smithery · Glama

6 tools · get_account_credits, get_compel

Eagle MCP Server is a local Model Context Protocol bridge that lets AI assistants (e.g. Cursor, Claude) access your Eagle media library so you can tag, categorize, search, and batch-organize assets with natural language.

Inkrun turns Markdown into consistently themed PDFs. Reach the render pipeline two ways — a REST API, or an MCP server your agent calls directly — plus a dashboard for authoring templates and themes. Add Inkrun as a custom connector in Claude Desktop over OAuth — no API key. ](https://inkrun.dev/docs/mcp/claude-desktop)[ Add Inkrun as a custom connector on claude.ai over OAuth — no API key. ](https://inkrun.dev/docs/mcp/claude-ai)[ Connect any other MCP client — over OAuth, or with an API key sent as a header.

5 tools · Extract structured data, Extract tables from PDFs

The Lottie Creator MCP brings your AI assistant directly into your animation workflow — giving it full access to LottieFiles Creator so it can build and edit Lottie animations on your behalf through natural language. You can check out the installation further down. MCP servers are part of a standardized interface for AI agents to interact with data sources using the Model Context Protocol. The Lottie Creator MCP connects AI assistants like Claude, Cursor, and Windsurf to the Creator API through a local bridge — meaning your AI gets the same full access to Creator that you do.

5 tools · Analyze for accessibility and brand, Batch generate animation variants

The Picsart MCP server exposes the full model catalog as Model Context Protocol tools. Connect it to any MCP-compatible agent and that agent can generate image, video, and audio across 178 models using natural language or structured tool calls. New to MCP? Start with What is MCP? first. 1. Install the gen-ai CLI — see Installation. 2. Run gen-ai login once (opens your browser for OAuth). That is all. The MCP server (gen-ai-mcp) ships with the CLI and uses the same credentials. > "Generate a product image on a white background using Flux 2 Pro, 4:3 aspect ratio."

5 tools · picsart_drive, picsart_generate

> Note > > MCP integration is currently available only in the desktop Editor for Windows and macOS. You can connect the Rive Editor to AI tools through MCP (Model Context Protocol). The first set of tools are designed to let AI handle repetitive tasks, like creating complex View Models, State Machines with hundreds of states/layers, Layouts, Shapes, and more. Create and manage Rive files — add, rename, resize, arrange, and focus artboards. Inspect and edit the scene — query hierarchies, select objects, update properties, rename, duplicate, reorder, reparent, or delete elements. Build designs — create shapes, paths, layouts, component instances, component lists, and asset-based elements. Create animation and interaction — edit linear animations, state machines, states, transitions, conditions, keyframes, and interpolation. Work with data binding — create view models, properties, instances, bindings, and custom property groups. Edit scripts and shaders — manage Luau scripts and WGSL shad

6 tools · Build designs, Create and manage Rive files

Screenies has a remote MCP server. Connect it to Claude Code, Claude.ai, Cursor, or any agent that speaks MCP, and the whole workflow runs from a prompt: upload your app screenshots, generate a polished set, localize it, and download files sized for App Store Connect. Create an account → Included with every account. Generation uses your normal credits. Then run /mcp in a session, pick screenies, and sign in. In Claude.ai, Cursor, and other MCP clients, add this URL as a remote MCP server (claude.ai: Settings, then Connectors, then Add custom connector):

6 tools · create_app, generate_screenshot_set

The canonical spec and reference implementation for the four components that raise a basket: quantity breaks, the free-shipping threshold bar, the order bump, and the sticky add-to-cart. MIT. No dependencies beyond React. Copy the file, or install the plugin and let your agent do it. On 25 August 2026 we asked ChatGPT and Perplexity, both with web search enabled, for a React component for buy 2, save 15%. ChatGPT returned a quantity stepper and cited no source at all. Zero annotations. Perplexity returned a stepper too, with min, max, step and a bundleSize prop, citing generic quantity-selector patterns.

5 tools · Add a free-shipping bar, Apply EU-compliant pricing rules

Wireflow speaks the Model Context Protocol. Connect Claude or any MCP client and let it build, inspect, and run your AI workflows, and generate and edit media through one canvas, API, and MCP, across 100+ models. It acts on your behalf with explicit consent and per-scope permissions. Add a custom connector in your MCP client and point it at the Streamable HTTP endpoint below. The client walks you through Wireflow’s OAuth sign-in and a consent screen, no manual API keys, no client setup. - OAuth 2.1 authorization code flow with PKCE (S256). - Dynamic Client Registration (RFC 7591), so clients self-register with no pre-shared credentials. - You approve a consent screen listing exactly the permissions the client asked for. Revoke anytime from settings.

6 tools · create_workflow_from_prompt, get_video_timeline

MCP server that turns local projects or GitHub repositories into Mermaid diagrams and renders them via Kroki.

Bidirectional Figma server over a local WebSocket relay: turn a selection into framework-aware code (reusing your components and tokens), or author frames, text, styles, variables, and whole screens back onto the canvas. 92 tools, any MCP client, no Dev Mode seat or paid tier.

Construction plan takeoff on stdio — load a plan PDF, read the title block, adopt the drawing scale (never applied silently), one-click trace rooms, and export real-world quantities (square footage, linear feet, counts) with provenance receipts. Install via NPM: `npx -y opentakeoff-mcp`

Turn any repo into an architecture diagram: deterministic extraction (file tree, JS/TS + Python import graph, frameworks, entry points) that the connected model refines into Mermaid or Graphviz DOT. Doubles as a CLI + GitHub Action enforcing architecture rules (forbidden imports, dependency cycles) with baselines in CI. Install via NPM: `npx -y repo-cartographer`

AI-powered Mermaid diagram generation with 22+ diagram types including flowcharts, sequence diagrams, class diagrams, ER diagrams, architecture diagrams, state machines, and more. Features 50+ pre-built templates, advanced layout engines, SVG/PNG/PDF exports, and seamless integration with GitHub Copilot, Claude, and any MCP-compatible client. Install via NPM: `npm install -g @narasimhaponnada/mermaid-mcp-server`

Generate beautiful Excalidraw architecture diagrams with auto-layout, architecture-aware component styling, and stateful editing. 50+ technology mappings including databases, message queues, caches, and more. No API keys required.

Agent-native UI/UX design validation: model user journeys as state machines and prove what AI-generated screens are missing (unreachable states, missing empty/loading/error states, WCAG contrast, touch targets, localization overflow) before code exists — then verify the code implements the contract. Deterministic CI gates (`uxloom check`, `uxloom audit`), one-command setup (`npx uxloom init`).

AI-to-CATIA Magic bridge for live SysML v1 models: 130+ MCP tools to query, audit, trace, and edit blocks, requirements, allocations, and diagrams in CATIA Magic Systems of Systems Architect 2026x over a local, air-gapped connection. FREE tier reads and audits; PRO/ENTERPRISE licences unlock writes and scripting.

Create branded images and PDFs for your campaigns and clients. Automate each new version through our API or your AI tools.

Art-directed website prompts as slash commands: hero sections, landing pages, SaaS, portfolio, e-commerce and 3D/WebGL scenes. Every build is registered as a real MCP prompt, so it lands in the command palette instead of waiting for the model to think of it, and is written natively for the tool you are in (Lovable, Bolt, Cursor or Claude). Each one has a live demo you can scroll before spending a token. Free prompts need no account; sign in from the session to unlock the rest. Install via NPM: `npx -y nightmarquee`

A Model Context Protocol (MCP) server for validating Mermaid diagram syntax and providing comprehensive grammar checking capabilities

Generate premium iOS mobile UI mockups (PNG + HTML) from a single text prompt. Pairs Claude (Opus 4.7 by default) with a Playwright headless renderer. Two backends — `claude` CLI (uses your Claude Code subscription) or Anthropic API. Stdio + HTTP transports, MIT.

Architectural decision tracking with testable predictions and in-band outcome gates for AI agents and engineering teams. Append-only JSONL storage with zero runtime dependencies. Install via PyPI: `pip install mcp-server-decisions`.

Curated Korean AEC knowledge — not just code lookup. Standards (KDS · KCS · KS · statutes) and domain expertise sit on one graph, so it answers **why** a figure holds, **what order** work goes in, and **how** two methods differ — each with the clause or the evidence attached. Abstains when it has no grounds. Remote endpoint `https://mcp.den.archi/mcp`, free beta at https://den.archi.

Autonomous cross-source inconsistency & contradiction intelligence engine: audits declared assertions across Kubernetes manifests, READMEs, OpenAPI specs, and Word documents to detect architectural drift. 12 canonical tools, deterministic scoring, and resolution advisory. Install: `npx -y contradiction-mcp`

Local design engine for AI agents. Create, edit and export real, editable designs (social, flyers, decks, business cards) with your brand applied; print-ready PDF with CMYK and bleed; imports InDesign, PSD and PDF. Powered by CE.SDK, no account needed. `npx -y @imgly/codesign-mcp@latest stdio`

Local MCP (`av_guard`) + CLI architecture gate: after AI edits, diff vs git HEAD for new cross-layer deps / structural changes (Apache-2.0; `npx arch-viewer setup`)

A complete color workflow over MCP: mix (incl. Kubelka-Munk paint physics), convert formats, generate harmonies, and remember named palettes. Resolves hex, CSS, and custom brand color names, and refuses unknown colors rather than guessing. Hosted, no install: `https://rnvizion-rnv-color-mcp.hf.space/mcp`.

AI-native architecture platform for engineering teams. Describe a system in natural language (e.g. "payment service with Stripe, Kafka, PostgreSQL") → get an interactive typed diagram with 167 cloud components → export to 14 production frameworks (Java Spring Boot, Python FastAPI, Go chi, Rust Axum, .NET, Kotlin Ktor, Ruby Rails, Elixir Phoenix, Docker Compose, Terraform, Mermaid, ADR, and more). Drift detection (`sync_architecture`) scores saved diagrams against live code 0–100 with a structured diff. 9 quality-lint rules catch orphans, SPOFs, god services. Also includes a secondary persistent code-graph layer for AI coding agents (free offline via `npx tentra-mcp --local init`). Agent-as-LLM pattern — zero LLM cost on our side, zero API key on yours. 35 MCP tools. Works in Cursor, Claude Code, Codex, Windsurf.

Let your AI author Mermaid diagrams (flowchart, sequence, ER, mindmap) in a shared workspace, then read, comment on, embed, and present them at a clean URL — instead of pasting back hundreds of lines of prose. Hosted endpoint `https://mcp.diagramzu.ai/mcp` or local `npx -y @diagramzu/mcp`. 14 tools incl. diagram CRUD, version history, comments, presentation decks, and `analyze_diagram` lints.

Local-first MCP server for AI-driven network design. Creates Cisco-style network master files, L1/L2/L3 topology diagrams, combined HTML viewers, device tables, and AI Context files.

Create workflow, sequence, and architecture diagrams using FlowZap Code DSL. 7 tools for validation, playground URL generation, syntax docs, graph export, artifact parsing, diffing, and patching. No API key required. Install via NPM: `npx -y flowzap-mcp`

Expose a company's complete design system (colors, typography, logos, components, guidelines, CSS tokens) to AI tools via MCP. Auto-parses CSS / Markdown / PDF / SVG / fonts; supports `marketing` vs `product` contexts with shared overrides; ships 12 tools, 16+ resources under `brandkit://`, and 4 prompts; stdio + SSE + Streamable HTTP transports. Install via NPM: `npx -y brandkit-mcp serve`.

"Mermaid for human movement" — a text language any LLM can write, rendered as a safe, anatomically-clamped 3D human figure in the browser.

Expert design & marketing knowledge for web / iOS / Android / macOS UI/UX: 83 sourced, prescriptive docs plus deterministic generators (design tokens, color systems, type scales, WCAG a11y) and a one-call `create_design_system`. Offline, no API key, reads only local files. Install via NPM: `npx -y saglitzdesign-mcp`.

Deterministic AI data-center design engine exposed as MCP tools (design, validate, layout): rack count, design PUE, total MVA, liquid/air cooling split, CDU planning, cost & timeline. NVIDIA Rubin-era, 22.9 kV intake, Korea. Remote Streamable HTTP at https://aidc-ai.io/api/mcp; no key for the anonymous tier.

MCP server for AI Diagram Maker — generate flowcharts, sequence diagrams, ERDs, system/network architecture, UML, mindmap, and workflow from natural language, code, ASCII, images, or Mermaid. Inline rendering using MCP Apps UI and editable diagram URLs. Requires API key.

5 tools · generate_diagram_from_ascii, generate_diagram_from_image

Catch bad animations before they ship. 13 deterministic checks on the live page (duration, easing, stagger, exits, reduced-motion, touch-gated hover) with no API key or LLM, plus optional vision-LLM UX review across viewports. Install via NPM: `npx -y motionlint mcp`