MCP server, CLI and agent skill for website traffic lookups with Vaneform (data: Similarweb estimates)
- ✓Open-source license (MIT)
- ✓Actively maintained (<30d)
- ✓Clear description
- ✓Topics declared
- ✓Documented (README)
claude mcp add vaneform-mcp -- npx -y github{
"mcpServers": {
"vaneform-mcp": {
"command": "npx",
"args": ["-y", "github"],
"env": {
"VANEFORM_API_KEY": "<vaneform_api_key>",
"VANEFORM_MCP_URL": "<vaneform_mcp_url>"
}
}
}
}VANEFORM_API_KEYVANEFORM_MCP_URLResumen de MCP Servers
# vaneform-mcp
[](https://github.com/jiangqizheng/vaneform-mcp/actions/workflows/ci.yml)
[](LICENSE)
Website traffic lookups for AI assistants and the terminal. This repo has an MCP server, a small CLI and an agent skill for [Vaneform](https://vaneform.com), a free website traffic checker.
Ask your assistant "how much traffic does notion.so get?" or "compare figma.com and canva.com" and it answers with estimated monthly visits, month-over-month change, rank, top countries, traffic sources and registry facts. Traffic figures are **Similarweb estimates**; see the [Vaneform methodology](https://vaneform.com/methodology).
[中文说明](README.zh-CN.md)
## What's inside
| Piece | What it is |
| --- | --- |
| Hosted MCP server | `https://vaneform.com/mcp` (Streamable HTTP). Zero install, and the recommended option. |
| `vaneform-mcp` | Local stdio MCP server for clients that only run local commands. It proxies the hosted server, so tools never drift. |
| `vaneform` CLI | `vaneform notion.so`, `vaneform compare a.com b.com`, `vaneform tld name`. It shares the same client. |
| [`skills/website-traffic-lookup`](skills/website-traffic-lookup/SKILL.md) | Agent skill (Claude / Agent Skills format) that tells an assistant when to check traffic and how to read and cite the result. |
## Tools
These tool names and descriptions come from the hosted server's `tools/list`:
| Tool | What it does | Cost |
| --- | --- | --- |
| `lookup_domain` | Cached snapshot of a domain: visits (full-site or search), registry, site profile, link-rank popularity | 0 points |
| `compare_bulk` | Side-by-side visits and registry for several domains; fetches traffic when it is missing | 0 points |
| `compare_tld` | One name across suffixes (default `com,ai,io`) | 0 points |
| `get_account` | Plan and remaining points | free |
| `lookup_keyword` | US keyword brief: volume, difficulty, CPC, Google occupancy (Pro) | 4 points on a cache miss |
## API key
Discovery (`initialize`, `tools/list`) is open. **Tool calls need a Vaneform API key.**
1. The API is included with Vaneform Pro. Create a key at [vaneform.com/account/api](https://vaneform.com/account/api). Domain lookups and compares cost no points; each key allows 60 requests a minute and 5,000 a day (429 `lookup_rate_exceeded` above that). Pro's 1,000 monthly points pay only for keyword data (`lookup_keyword`, keyword rankings).
2. Pass the key as `Authorization: Bearer vf_live_…` (remote) or `VANEFORM_API_KEY` (local server and CLI).
If no key is set, tools still list. A call returns a sign-up link and a browser link such as `https://vaneform.com/?q=notion.so`, where you get 3 free lookups a day without an account (999 with a free account). When points run out, the error points to [pricing](https://vaneform.com/pricing).
## Install
### Claude Code (remote)
```bash
claude mcp add --transport http --scope user vaneform https://vaneform.com/mcp \
--header "Authorization: Bearer vf_live_…"
```
### Cursor (remote)
`~/.cursor/mcp.json` (or `.cursor/mcp.json` in a project):
```json
{
"mcpServers": {
"vaneform": {
"url": "https://vaneform.com/mcp",
"headers": { "Authorization": "Bearer vf_live_…" }
}
}
}
```
### Claude Desktop (local stdio)
The simplest Claude Desktop setup that passes your key is the local proxy. Edit `claude_desktop_config.json` (macOS: `~/Library/Application Support/Claude/`, Windows: `%APPDATA%\Claude\`):
```json
{
"mcpServers": {
"vaneform": {
"command": "npx",
"args": ["-y", "github:jiangqizheng/vaneform-mcp"],
"env": { "VANEFORM_API_KEY": "vf_live_…" }
}
}
}
```
The package isn't on npm yet, so `npx` installs it from GitHub. The first start takes a little longer because it builds once. If `npx` exits right away on an old system npm (Debian's packaged npm 9.2 does this with GitHub packages), run `npm install -g github:jiangqizheng/vaneform-mcp` and use `"command": "vaneform-mcp"` with no args.
### Codex
```bash
export VANEFORM_API_KEY=vf_live_…
codex mcp add vaneform --url https://vaneform.com/mcp --bearer-token-env-var VANEFORM_API_KEY
```
### Any other MCP client
- Remote: URL `https://vaneform.com/mcp`, transport Streamable HTTP, header `Authorization: Bearer vf_live_…`
- Local: command `npx -y github:jiangqizheng/vaneform-mcp`, env `VANEFORM_API_KEY`
## CLI
```bash
npx -p github:jiangqizheng/vaneform-mcp vaneform --help
# or clone, then: npm install && npm link
export VANEFORM_API_KEY=vf_live_…
vaneform notion.so # traffic overview
vaneform compare notion.so coda.io # side by side
vaneform tld vaneform --suffixes com,ai,io
vaneform account # remaining points (free)
vaneform notion.so --json # raw API JSON
```
Real output against production on 2026-10-08, with no key set:
```text
$ vaneform tools
lookup_domain Read a cached Vaneform domain snapshot: scale (full-site or search visits), registry, site profile, and link-rank popularity.
get_account Read the caller plan, website lookup allowance, and point balance.
lookup_keyword Read a US/EN keyword brief: monthly searches, difficulty, CPC, and Google occupancy.
compare_bulk Compare domains side by side.
compare_tld Compare the same registrable name across suffixes.
$ vaneform github.com
VANEFORM_API_KEY is not set, so this tool cannot call the Vaneform API.
API access is included with Pro. Create an API key at https://vaneform.com/account/api (domain lookups cost no points).
Check it in a browser instead: https://vaneform.com/?q=github.com (3 free lookups a day without an account, 999 with a free account).
```
With a key, `vaneform <domain>` prints visits, MoM change, rank, top countries, sources, registry dates and the report link, followed by `Data: Vaneform (data: Similarweb estimates)`.
## Talk to the hosted server directly
```bash
curl -s https://vaneform.com/mcp \
-H 'content-type: application/json' -H 'accept: application/json, text/event-stream' \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'
```
The same key works for the REST API: [docs](https://vaneform.com/docs/api), [OpenAPI](https://vaneform.com/api/v1/openapi.json).
## Agent skill
Copy the skill so your assistant knows when to reach for traffic data and how to cite it:
```bash
# Claude Code (user-wide)
mkdir -p ~/.claude/skills/website-traffic-lookup
curl -fsSL https://raw.githubusercontent.com/jiangqizheng/vaneform-mcp/main/skills/website-traffic-lookup/SKILL.md \
-o ~/.claude/skills/website-traffic-lookup/SKILL.md
```
For other agents that read `SKILL.md` folders, put it in that agent's skills directory (for example `.agents/skills/` in a project). Vaneform also serves a lower-level API skill at [vaneform.com/skill.md](https://vaneform.com/skill.md).
## Reading the numbers
- `status: "missing"` means Vaneform hasn't cached that domain yet. It does not mean zero traffic. `compare_bulk` fetches it.
- Every estimate carries its month (`as_of`) and a `confidence` label. Quote them.
- Cite results as "Vaneform (data: Similarweb estimates)".
## Development
```bash
npm install # also builds dist/
npm run lint # biome + tsc
npm test # unit tests + a stdio test with the official MCP SDK client
VANEFORM_MCP_URL=http://localhost:3000/mcp node dist/bin/vaneform.js tools # point at another server
```
The runtime has no dependencies and needs Node 18 or newer. This repo contains only the client. No data collection code lives here.
## License
[MIT](LICENSE). The license covers this code only, not data returned by the API. Traffic figures are Similarweb estimates served by [Vaneform](https://vaneform.com).
Lo que la gente pregunta sobre vaneform-mcp
¿Qué es jiangqizheng/vaneform-mcp?
+
jiangqizheng/vaneform-mcp es mcp servers para el ecosistema de Claude AI. MCP server, CLI and agent skill for website traffic lookups with Vaneform (data: Similarweb estimates) Tiene 0 estrellas en GitHub y su última actualización registrada es del 2026-10-10.
¿Cómo se instala vaneform-mcp?
+
Puedes instalar vaneform-mcp clonando el repositorio (https://github.com/jiangqizheng/vaneform-mcp) o siguiendo las instrucciones del README en GitHub. ClaudeWave también te ofrece bloques de instalación rápida en esta misma página.
¿Es seguro usar jiangqizheng/vaneform-mcp?
+
Nuestro agente de seguridad ha analizado jiangqizheng/vaneform-mcp y le ha asignado un Trust Score de 95/100 (tier: Verified). Revisa el desglose completo de comprobaciones superadas y flags en esta página.
¿Quién mantiene jiangqizheng/vaneform-mcp?
+
jiangqizheng/vaneform-mcp es mantenido por jiangqizheng. La última actividad registrada en GitHub es del 2026-10-10, con 0 issues abiertos.
¿Hay alternativas a vaneform-mcp?
+
Sí. En ClaudeWave puedes explorar mcp servers similares en /categories/mcp, ordenados por popularidad o actividad reciente.
Despliega vaneform-mcp en tu cloud
Lleva este repo a producción en minutos. Cada plataforma genera su propio entorno con variables de entorno editables.
¿Mantienes este repo? Añade un badge a tu README
Pega el badge en tu README de GitHub para mostrar que está auditado por ClaudeWave. Cada badge enlaza de vuelta a esta página y muestra el Trust Score actual.
[](https://claudewave.com/repo/jiangqizheng-vaneform-mcp)<a href="https://claudewave.com/repo/jiangqizheng-vaneform-mcp"><img src="https://claudewave.com/api/badge/jiangqizheng-vaneform-mcp" alt="Featured on ClaudeWave: jiangqizheng/vaneform-mcp" width="320" height="64" /></a>Más MCP Servers
Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, self-host or cloud, 400+ integrations.
User-friendly AI Interface (Supports Ollama, OpenAI API, ...)
An open-source AI agent that brings the power of Gemini directly into your terminal.
Real-time global intelligence dashboard. AI-powered news aggregation, geopolitical monitoring, and infrastructure tracking in a unified situational awareness interface
🕷️ An adaptive Web Scraping framework that handles everything from a single request to a full-scale crawl! Don't be shy, join here: https://discord.gg/EMgGbDceNQ and follow here for daily tips and tricks: https://x.com/Scrapling_dev
The fastest path to AI-powered full stack observability, even for lean teams.