Zenith logo

Zenith

Creates, validates, edits, and deterministically renders .zen design documents to PNG, SVG, or PDF through MCP.

Install Zenith

Pick your client, then add it in one click or paste the config.

STDIONode.js
~/Library/Application Support/Claude/claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "zenith": {
      "command": "npx",
      "args": [
        "-y",
        "@zenitheditor/zenith-mcp"
      ]
    }
  }
}

Settings → Developer → Edit Config, merge this into mcpServers, then fully quit and reopen Claude Desktop.

Install not yet sandbox-tested— we haven't run this command through our automated sandbox yet. Not a red flag; we're working through the catalog.

Overview

zenith MCP server brings Zenith’s plain-text .zen design-document engine to AI agents for authoring, validation, editing, inspection, and rendering. It processes human-readable KDL-based files, reports diagnostics, and compiles documents into deterministic PNG, self-contained SVG, or print-ready PDF output. Reach for zenith when design files need to be reviewed, diffed, safely modified, or rendered consistently in an agent workflow. The server is distributed through the @zenitheditor/zenith-mcp npm package, which installs a matching prebuilt Zenith binary.

Use cases

  • Author .zen design documents with an AI agent
  • Validate and automatically fix design source
  • Inspect tokens and document node trees
  • Render multipage designs to PNG, SVG, or PDF
  • Review deterministic design changes in version control

Key features

  • Plain-text KDL-based .zen documents
  • Structured validation and machine fixes
  • Canonical formatting and node inspection
  • Deterministic PNG, SVG, and PDF rendering
  • Multipage export and scene JSON output
  • Rasterization diagnostics and export policies

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked 3d ago

Recent health check succeeded.

8/8 checks healthy over the last 50d

Supply-chain signal

No high-severity advisories surfaced by our automated scan.

Critical 0High 0Medium 0Low 0

Scanned 12d ago via OSV.dev · @zenitheditor/zenith-mcp (npm)

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

GitHub stars22Stargazers on the source repository.
npm downloads286Package downloads in the last 30 days.
Last commit4d agoMost recent push to the default branch.
Directory activity2 viewsConfig copies, upvotes, and views on AllMCPs.

How Zenith works

What zenith does

zenith provides an MCP interface to Zenith, a design-document format and rendering engine built around plain-text .zen files. These files use human-readable KDL syntax and can represent posters, presentations, books, social graphics, diagrams, and other designs. Because the source is text, it can be read, reviewed, diffed, validated, and edited by both developers and AI agents.

The engine turns a .zen document into a backend-neutral scene and renders that scene deterministically. The same source produces the same pixels on repeated runs, making the format suitable for workflows where reproducibility matters. Output formats include PNG, self-contained SVG, and print-ready PDF.

How it works

Zenith parses a .zen file, validates it against its diagnostic system, and then compiles it for inspection or export. Validation can produce regular or JSON-formatted diagnostics. Formatting produces canonical, idempotent source text, while machine fixes can apply supported corrections such as token or typo fixes.

Rendering supports single-page and multipage documents. PNG output can be scaled, SVG output can target a selected page or write one file per page, and PDF output can contain every page or a selected page. The engine can also emit its scene intermediate representation as JSON. Export reports identify rasterized regions and explain why fallback rendering was required.

Setup and configuration

The MCP-focused installation uses npm:

Terminal
npm install -g @zenitheditor/zenith-mcp
npx -y @zenitheditor/zenith-mcp --help

The package installs a matching prebuilt zenith binary from GitHub Releases. Linux, macOS, and Windows installation paths are supported. The underlying CLI can also be installed with the platform install scripts, Cargo, or a release archive. Cargo users can install zenith-tool, which provides the zenith executable.

For script-based installation, ZENITH_INSTALL_DIR changes the destination directory; the default is ~/.local/bin. Source builds use Cargo and can install the release binary from the repository workspace.

Tools and capabilities

Agent workflows can use Zenith operations for:

  • Creating and editing .zen design documents.
  • Validating source and requesting machine-readable diagnostics.
  • Applying supported automatic fixes.
  • Formatting source into canonical KDL text.
  • Listing design tokens and resolved values.
  • Inspecting the document node tree without modifying it.
  • Rendering PNG, SVG, and PDF files.
  • Rendering all pages or selecting an individual page.
  • Exporting the compiled scene as JSON.

PNG output accepts a scale value from greater than zero through 4. SVG and PDF fallback resolution can be adjusted separately with --raster-scale. Export policies can deny SVG or PDF raster fallback, and blocking diagnostics prevent output writes. SVG text is outlined, so it preserves appearance but loses text selection and search. Captured PDF content can similarly lose text selection, search, or link targets.

Limitations and notes

Some visual features require raster fallback during SVG or PDF export. Reports identify affected pages, command ranges, and reasons. Existing bitmap assets retain their source resolution, while --raster-scale changes fallback pixels without changing vector geometry or page dimensions.

File writes use a temporary sibling followed by a rename, preserving an existing destination when a write fails. In batch operations, earlier committed outputs may remain after a later I/O error, and result rows retain committed paths alongside failure status. Read-only destinations are rejected.

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about Zenith

zenith is an MCP server that connects AI agents to Zenith’s plain-text .zen design-document engine. Its main tools support authoring, validating, editing, formatting, inspecting, and rendering design files to deterministic PNG, self-contained SVG, and print-ready PDF outputs.

Install it with `npm install -g @zenitheditor/zenith-mcp`. The package installs a matching prebuilt `zenith` binary from GitHub Releases, and you can verify the command with `npx -y @zenitheditor/zenith-mcp --help`.

zenith renders `.zen` design documents to PNG, self-contained SVG, and print-ready PDF. It supports multipage output, individual page selection, one SVG per page, and scene intermediate-representation output as JSON.

Yes, zenith validates `.zen` files and can apply supported machine fixes. Validation reports diagnostics, optionally as JSON, while the fix operation can apply corrections such as token and typo fixes.

Yes, zenith is designed to render the same `.zen` source to the same pixels on every run. Its engine compiles the document to a backend-neutral scene before producing PNG, SVG, or PDF output.

Related MCP servers

Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG PreviewZenith AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/zenith?style=directory)](https://allmcps.com/mcp/zenith)
HTML Embed
<a href="https://allmcps.com/mcp/zenith"><img src="https://allmcps.com/api/badge/zenith?style=directory" alt="Zenith on AllMCPs" /></a>