# AIGUI (npm · @ai-gui/mcp)

Let coding agents draw charts, diagrams, 3D, molecules and animated topologies as PNGs or live pages

- Trust score: 74/100 (medium)
- Change this week: +39
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-10-23

## Components

- npm · `@ai-gui/mcp`: 74/100 (this document), [markdown](https://verifymcp.io/servers/liliang-cn-aigui/ai-gui-mcp.md), [page](https://verifymcp.io/servers/liliang-cn-aigui/ai-gui-mcp)

## Channel facts

- Registry: `npm`
- Package: `@ai-gui/mcp`
- Version: `0.49.0`
- 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 2026-10-23.

- **Supply Chain Security**: 88/100
  - No malware found by supply-chain analysis.
  - CVE check failed: a known medium-severity CVE affects echarts 5.6.0, a direct dependency. A fixed version is available.
  - No install/post-install scripts declared.
  - 126 of 317 dependencies flagged as unhealthy.
- **Provenance & Transparency**: 45/100
  - Source repository is publicly reachable at the declared URL.
  - Provenance check failed: no build-provenance attestation is published.
  - Clear OSI-approved license (MIT).
  - Actively maintained (last published 12 days ago).
  - Disclosure check failed: no security disclosure policy was found in the source repository.
- **Schema Quality & AI Usability**: 66/100
  - AI-judged instruction clarity (excellent).
  - Context-footprint check failed: tool/resource definitions use about 1459 tokens (~208/item across 7 items; 7 tools + 0 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 63/100
  - Stability observed for 19 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 94/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 83% 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 7 captured tool definition(s), and no name or description among them implies an irreversible operation.
  - An AI judge read all 7 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 AIGUI MCP server?

AIGUI runs locally as an npm package, launched with npx -y @ai-gui/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 liliang-cn-aigui -- npx -y @ai-gui/mcp
```

### Cursor

```json
{
  "mcpServers": {
    "liliang-cn-aigui": {
      "command": "npx",
      "args": [
        "-y",
        "@ai-gui/mcp"
      ]
    }
  }
}
```

### VS Code

```json
{
  "servers": {
    "liliang-cn-aigui": {
      "command": "npx",
      "args": [
        "-y",
        "@ai-gui/mcp"
      ]
    }
  }
}
```

### Codex

```bash
codex mcp add liliang-cn-aigui -- npx -y @ai-gui/mcp
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "liliang-cn-aigui": {
      "type": "local",
      "command": [
        "npx",
        "-y",
        "@ai-gui/mcp"
      ],
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add liliang-cn-aigui --command npx --arg -y --arg @ai-gui/mcp
```

### Hermes

```yaml
mcp_servers:
  liliang-cn-aigui:
    command: "npx"
    args: ["-y", "@ai-gui/mcp"]
```

### Netclaw

```json
{
  "McpServers": {
    "liliang-cn-aigui": {
      "Transport": "stdio",
      "Command": "npx",
      "Arguments": [
        "-y",
        "@ai-gui/mcp"
      ]
    }
  }
}
```

### Vellum

```bash
assistant mcp add liliang-cn-aigui -t stdio -c npx -a -y @ai-gui/mcp
```

### Other

```json
{
  "mcpServers": {
    "liliang-cn-aigui": {
      "command": "npx",
      "args": [
        "-y",
        "@ai-gui/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.

### 2026-10-23 (score 74, +1)

No change was recorded against any check on this day. Stability & Change Management went from 60 to 63. That category is still filling its 30-day observation window: 18 days of observed history at the previous scan, 19 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-10-21 (score 73, +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.

### 2026-10-19 (score 72, +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.

### 2026-10-17 (score 71, +36)

- [security improvement] Injection markers: unverified → pass
- [functional improvement] Tool coverage: unverified → 100
- [functional improvement] Stability: unverified → 0.43
- [functional improvement] MCP protocol: unverified → pass
- [functional] Package version: 0.48.0 → 0.49.0

### 2026-10-16 (score 35, 0)

- [functional] Package version: 0.48.0 → 0.49.0

### 2026-10-14 (score 35, 0)

- [security] Stability: Stability not yet verified: this package could not be installed in a clean environment, so we have no schema to compare.
- [security] Tool safety: Tool safety not yet verified: this package could not be installed in a clean environment, so we have no tool text to scan.
- [functional] Schema quality: Schema quality not yet verified: this package could not be installed in a clean environment, so we have no schema to assess.
- [functional] Capabilities: Protocol version not yet verified: this package could not be installed in a clean environment, so we never saw its MCP handshake.
- [functional] Tool coverage: Tool coverage not yet verified: this package could not be installed in a clean environment, so we have no tool definitions to assess.
- [functional] Package version: 0.48.0 → 0.49.0

### 2026-10-13 (score 35, 0)

- [functional regression] Security disclosure: unverified → fail

### 2026-10-12 (score 35, −30)

- [security regression] Tool safety: pass → unverified
- [security] Stability: Stability not yet verified: we do not have a sandbox capture of the MCP schema this version of the package serves yet.
- [functional regression] Security disclosure: fail → unverified
- [functional regression] Capabilities: pass → unverified
- [functional regression] Tool coverage: 100 → unverified
- [functional] First check of Schema quality: unverified
- [functional] Package version: 0.48.0 → 0.49.0

## MCP tools (7)

### `aigui_guide` (~172 tokens)

How to write AIGUI blocks

Get the exact syntax for AIGUI blocks before writing any. Call with no arguments for the list of blocks, then with the names you will use. Blocks: chart, mermaid, katex, bigscreen, dashboard, scene, gravity, topology, molecule, graph, solid, function, optics, motion, physics, figure, quote, primitives, highlight, progress, citation. Each block is a fenced code block of JSON (or mermaid text, or $$ maths) inside ordinary markdown. The rules matter: most blocks take conditions, not results, and compute the numbers themselves.

Input parameters:

- `blocks` (array): Block names to get the syntax for, e.g. ["chart", "scene"]. Omit for the list.
- `locale` (string): Language of the guidance. Default en.

### `aigui_render` (~328 tokens)

Draw AIGUI blocks as PNG

Draw the AIGUI blocks in a markdown string as PNG images and return them. Draws: chart, mermaid, katex, bigscreen, dashboard, scene, gravity, topology, molecule, and markdown tables. Anything else in the markdown is ignored. Use it when a picture answers better than text — a chart of numbers, a diagram of a flow, a 3D shape, an orbit, a molecule. The images come back to you and are saved to disk; the result says whether they were also opened on the user's screen — if not, tell the user the paths. Call aigui_guide first for the syntax of any block you have not written in this session.

Input parameters:

- `data` (object): Local data files by name, e.g. {"sales": "/abs/path/sales.csv"} (.csv, .tsv, .json). In a block's JSON, {"$data":"sales"} becomes its rows, {"$data":"sales","column":"revenue"} one column, {"$data":"…
- `markdown` (string, required): Markdown containing one or more AIGUI blocks.
- `theme` (string)
- `width` (integer): Picture width in CSS pixels. Default 720; use ~1100 for a bigscreen.

### `aigui_open` (~362 tokens)

Open AIGUI blocks in the browser

Write a markdown answer with AIGUI blocks to an HTML page and open it in the user's browser, where charts are interactive, 3D scenes and molecules can be turned, and walls and orbits animate. Draws every block: chart, mermaid, katex, bigscreen, dashboard, scene, gravity, topology, molecule, graph, solid, function, optics, motion, physics, figure, quote, primitives, highlight, progress, citation, plus maths, tables and code. Use it for a long answer that is mostly visual, for anything the user will want to explore, or for a [page only] block. Returns the file path. Call aigui_guide first for the syntax of any block you have not written in this session.

Input parameters:

- `data` (object): Local data files by name, e.g. {"sales": "/abs/path/sales.csv"} (.csv, .tsv, .json). In a block's JSON, {"$data":"sales"} becomes its rows, {"$data":"sales","column":"revenue"} one column, {"$data":"…
- `locale` (string): Language of the page's own buttons and labels. Default en; use the user's language when they ask for one.
- `markdown` (string, required): The whole answer: prose and AIGUI blocks.
- `theme` (string): Default: follow the system setting.
- `title` (string): Page title.

### `aigui_edit` (~170 tokens)

Change a page aigui_open wrote

Change part of a page written by aigui_open — a number, a panel, a step — without sending the whole answer again. Each edit replaces text in the page's markdown: find must be quoted exactly from what you wrote and occur once; include enough of the surrounding JSON to make it unique. The page is rewritten in place, opened again, and looked over like a new one. page defaults to the most recent one.

Input parameters:

- `edits` (array, required): Find-and-replace edits on the page's markdown, applied in order.
- `locale` (string): Change the language of the page's own buttons and labels.
- `page` (string): The page file path or name, or "last" (default).
- `title` (string): A new title, if it should change.

### `aigui_topology` (~167 tokens)

Draw a topology from real config

Read a docker-compose file, a Kubernetes manifest, or a directory of manifests and draw the system it describes as a topology — services, databases, queues, Services and Ingresses, volumes, and what connects to what — from the config itself rather than from memory. Opens it as a page and returns the ```topology block, which you can extend with steps (a request's path, a failover) and draw again with aigui_open.

Input parameters:

- `locale` (string)
- `open` (boolean): Open it as a page (default true). false returns only the block.
- `path` (string, required): Absolute path to a compose file, a manifest, or a directory of .yaml manifests.
- `theme` (string)
- `title` (string)

### `aigui_feedback` (~91 tokens)

Read the reader's comments on pages

Read the comments the user left on pages opened by aigui_open — each names the page, the block and what they want changed. Call it when the user says they commented, or asks you to look at their notes. Comments are taken when read (keep: true leaves them). Act on them with aigui_edit.

Input parameters:

- `keep` (boolean): Leave the comments in place after reading. Default false.

### `aigui_export` (~169 tokens)

Save a page as PNG or PDF

Save a page written by aigui_open as one full-length PNG, a PDF, a single self-contained HTML file that stays interactive, or a GIF / WebM of it playing — for sending or attaching. Returns the file path. page defaults to the most recent one.

Input parameters:

- `format` (string): png (default) or pdf: drawn still. html: the page as one self-contained file that stays interactive — for sending to someone. gif or webm: one play of what moves on the page — a topology's or a scene…
- `page` (string): The page file path or name, or "last" (default).
- `theme` (string): Colour scheme to draw it in. Default light.

## Diagnostics

Captured diagnostic sections: Provenance, Vulnerabilities, Dependencies. The full working is on the page: https://verifymcp.io/servers/liliang-cn-aigui/ai-gui-mcp#diagnostics

## Score history

- 2026-10-23: 74
- 2026-10-22: 73
- 2026-10-21: 73
- 2026-10-20: 72
- 2026-10-19: 72
- 2026-10-18: 71
- 2026-10-17: 71
- 2026-10-16: 35
- 2026-10-15: 35
- 2026-10-14: 35
- 2026-10-13: 35
- 2026-10-12: 35
- 2026-10-08: 65
- 2026-10-05: 65
- 2026-10-04: 65

## Common questions

### What is the AIGUI MCP server?

AIGUI is an MCP server listed in the public MCP registry as io.github.liliang-cn/aigui. Let coding agents draw charts, diagrams, 3D, molecules and animated topologies as PNGs or live pages. This page covers its npm package (@ai-gui/mcp).

### Is the AIGUI MCP server safe to use?

AIGUI scores 74 out of 100 on VerifyMCP. We recorded 2 known advisories against it as of 23 October 2026. It declares no install or post-install scripts. 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 AIGUI MCP server expose?

AIGUI exposes 7 tools: aigui_guide, aigui_render, aigui_open, aigui_edit, aigui_topology, and 2 more. Their descriptions and schemas cost roughly 1,459 tokens of context every time the server is loaded.

### Is the AIGUI MCP server still maintained?

AIGUI is still listed as active in the MCP registry. We last reached this channel on 23 October 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 AIGUI MCP server under?

AIGUI declares the MIT licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.

## Links

- npm package: https://www.npmjs.com/package/@ai-gui/mcp
- Socket report: https://socket.dev/npm/package/@ai-gui/mcp
- Repository: https://github.com/liliang-cn/aigui
- Website: https://liliang-cn.github.io/aigui/
- Changelog RSS feed: https://verifymcp.io/servers/liliang-cn-aigui/ai-gui-mcp.xml
- Changelog JSON feed: https://verifymcp.io/servers/liliang-cn-aigui/ai-gui-mcp.json
- HTML version of this page: https://verifymcp.io/servers/liliang-cn-aigui/ai-gui-mcp