Napkin

io.github.mindstonePublished by GitHub account mindstonev0.4.1Updated Oct 25, 2026

Napkin AI visual generation MCP server for Model Context Protocol hosts

VerifiedSTDIODesktop onlyMedia & DesignAI & ML

Overview

AI-generated overview

Napkin is a local MCP server that lets an assistant generate Napkin AI visuals such as diagrams, infographics and illustrations from text, check generation jobs

What you can do
Generate a professional visual from a text description, optionally choosing one of the built-in styles and passing context text. Check the status of a generation request. Download a finished visual file to disk without overwriting existing files. List the 15 built-in visual styles, which needs no API key or network call. Configure the Napkin AI API key from within the session.
When to use it
Useful when you want an assistant to turn written explanations into diagrams, infographics or illustrations and save the results locally, rather than describing a visual in words. Suited to desktop MCP hosts where a Napkin AI API key is available.
Requirements
Runs locally via npx with Node.js 20+ and npm. Requires a Napkin AI API key in NAPKIN_API_KEY and network access to the Napkin API. Optional settings: MCP_WORKSPACE_PATH for where generated visuals are saved, and NAPKIN_REQUEST_TIMEOUT_MS for the outbound request timeout (default 60 seconds, maximum 30 minutes).
Before you install
The Napkin API is a developer preview, so endpoints and behavior may change. The API key is a secret passed by environment variable; the server can also configure it during a session. Generation and download calls send your text prompts to Napkin's cloud service. Downloads write files to disk, though existing files are never overwritten.

Installation

In SourceWeft

  1. Open Napkin in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

@mindstone/mcp-server-napkin

[npm version] [License: FSL-1.1-MIT]

Napkin AI visual generation MCP server for Model Context Protocol hosts. Generate professional visuals — diagrams, infographics, and illustrations — from text descriptions, check generation status, and download results through a standardised MCP interface.

Status

Note: The Napkin API is currently a developer preview (see api.napkin.ai). Endpoints and behavior may change; report upstream issues to [email protected]. Older guides reference context_before/context_after parameters — Napkin deprecated these in favor of the single context parameter, which napkin_generate_visual supports.

Requirements

  • Node.js 20+
  • npm

One-click install

[Add to Cursor] [Add to VS Code] [Add to VS Code Insiders]

After clicking the button, your host will prompt you to fill: NAPKIN_API_KEY, NAPKIN_REQUEST_TIMEOUT_MS.

Manual config for Claude Desktop / Claude Code / Goose / Continue.dev (Napkin)
json
{  "mcpServers": {    "Napkin": {      "command": "npx",      "args": [        "-y",        "@mindstone/mcp-server-napkin"      ],      "env": {        "NAPKIN_API_KEY": "",        "NAPKIN_REQUEST_TIMEOUT_MS": "60000"      }    }  }}

Quick Start

Install & build

bash
cd <path-to-repo>/connectors/napkinnpm installnpm run build

npx (once published)

bash
npx -y @mindstone/mcp-server-napkin

Local

bash
node dist/index.js

Configuration

Environment variables

  • NAPKIN_API_KEY — Napkin AI API key
  • MCP_WORKSPACE_PATH — optional workspace path for saving generated visuals
  • MCP_HOST_BRIDGE_STATE — optional path to a host bridge state file used for credential management
  • MINDSTONE_REBEL_BRIDGE_STATE — backwards-compatible alias for MCP_HOST_BRIDGE_STATE
  • NAPKIN_REQUEST_TIMEOUT_MS — optional override (positive integer ms, max 30 min) for the outbound HTTP request timeout applied to Napkin API, host-bridge, and download calls. Default: 60000 (60s). Raise this if you see TIMEOUT errors on slow submits or large downloads; lower it if you want tighter bounds.

Host configuration examples

Claude Desktop / Cursor

json
{  "mcpServers": {    "Napkin": {      "command": "npx",      "args": ["-y", "@mindstone/mcp-server-napkin"],      "env": {        "NAPKIN_API_KEY": "your-api-key"      }    }  }}

Local development (no npm publish needed)

json
{  "mcpServers": {    "Napkin": {      "command": "node",      "args": ["<path-to-repo>/connectors/napkin/dist/index.js"],      "env": {        "NAPKIN_API_KEY": "your-api-key"      }    }  }}

Tools (5)

Configuration

  • configure_napkin_api_key — Configure the Napkin AI API key

Generation

  • napkin_generate_visual — Generate a professional visual from text
  • napkin_check_status — Check the status of a visual generation request (vendor-authored free text — warning/error messages and visual_query — is wrapped in <untrusted-content> envelopes so hosts treat it as data; codes, IDs, and URLs stay raw, with their machine-identifier charset enforced by response validation)
  • napkin_download_visual — Download a generated visual file to disk (existing files are never overwritten — pass a different filename or omit it to auto-generate one). Redirects are re-validated against the Napkin host allow-list before being followed; off-list redirect targets are refused.

Styles

  • napkin_list_styles — List the 15 built-in Napkin visual styles (IDs, descriptions, categories) to pick a style_id for napkin_generate_visual. Static catalog; no API key or network call required.

Licence

FSL-1.1-MIT — Functional Source License, Version 1.1, with MIT future licence. The software converts to MIT licence on 2030-04-08.

Source: connectors/napkin/README.md at commit 22c4611

Tools

0
SourceWeft has not started this package in its sandbox yet, so its tools are not known.

Version history

1
  1. v0.4.1LatestSep 16, 2026

Score

40% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 days25 pts

    Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.

  • Auth declared10 pts

    Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.

  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts

    Declare a license in the repository or the registry entry.