100Critical
The skill allows arbitrary code execution via its
Behavior Analysis
Description verified — behavior matches claim
Claims to do
@json-render/mcp: MCP Apps integration that serves json-render UIs as interactive MCP Apps inside Claude, ChatGPT, Cursor, VS Code, and other MCP-capable clients.
Actually does
This skill provides a Node.js server implementation using `@json-render/mcp` to create an MCP application. It defines a catalog of UI components, registers a `render-ui` tool, and serves an HTML resource by reading `dist/index.html` from the filesystem. Client-side React code uses `useJsonRenderApp` to render UIs received via `postMessage` within an iframe, with client configurations provided to execute the Node.js server via `npx tsx server.ts --stdio`.
Skill Info
Namevercel-labs/json-render/mcp
Registrygithub
Version14873b8de451
PURLpkg:github/vercel-labs/json-render@14873b8de451?skill=mcp
Stars14,582
Installs573
Source
SHA-256706ccd32b510...
SHA-1ababd91a0a8e...
MD5f06c53958fec...
TLSH837182515523...
Size3,528 bytes
Install
This skill has been flagged as potentially malicious. Review the findings below before installing.
Assessments (1)
AI Agent Traps ↗Arbitrary Command Execution via MCP Server Configcriticalcommand execution
The skill's configuration for MCP clients (Cursor, Claude Desktop) explicitly defines a 'command' field that executes arbitrary shell commands (e.g., 'npx tsx server.ts'). This allows for arbitrary code execution on the host system if the configuration is compromised or maliciously crafted.
100% confidence
command: "npx", args: ["tsx", "server.ts", "--stdio"]
Local File System Read Accesslowdata exfiltration
The server component reads local files (e.g., 'dist/index.html') using 'fs.readFileSync'. While used for legitimate purposes in the example, this capability could be exploited by a compromised server to read sensitive files from the local filesystem.
70% confidenceline 31
html: fs.readFileSync("dist/index.html", "utf-8")Badge
Markdown
[](https://mondoo.com/ai-agent-security/skills/github/vercel-labs/json-render/mcp)HTML
<a href="https://mondoo.com/ai-agent-security/skills/github/vercel-labs/json-render/mcp"><img src="https://mondoo.com/ai-agent-security/api/badge/github/vercel-labs/json-render/mcp.svg" alt="Mondoo Skill Check" /></a>Image URL
https://mondoo.com/ai-agent-security/api/badge/github/vercel-labs/json-render/mcp.svgSecure your AI agents
Skills can read files, run commands, and access credentials. Mondoo helps organizations manage the security risks of AI agent skills across their entire fleet.
- Continuous scanning of skills across all registries
- Policy enforcement before skills reach your agents
- Integration with your existing security workflow