Skip to main content
Glama
Vinted MCP and CLI Server

Vinted MCP and CLI Server

🛍️ Vinted MCP & CLI Server

Gib deinem KI-Assistenten Zugriff auf Vinted — durchsuche Artikel, vergleiche Preise und verfolge Verkäufer in 19 Ländern.

npm version CI License: MIT Node ≥18


Die Idee

Vinted hat keine öffentliche API. Dieses Paket schließt diese Lücke — es ermöglicht KI-Assistenten, direkt mit Vinted zu kommunizieren über das Model Context Protocol.

Verbinde es mit Claude, Cursor oder einem anderen MCP-kompatiblen Assistenten und frage einfach:

"Suche mir eine North Face Jacke unter 60 € in Deutschland, Zustand gut oder besser"

"Vergleiche die Preise für Air Jordan 1 in Frankreich, Italien und Großbritannien"

"Was verkauft Verkäufer #123456 gerade? Irgendetwas unter 20 €?"

Die KI findet heraus, welche Filter zu verwenden sind, ruft Vinted auf und gibt dir eine echte Antwort — kein Suchen, kein Filtern, keine Tabs.

Ebenfalls als CLI-Tool und TypeScript-Bibliothek für die direkte Verwendung verfügbar.


Related MCP server: qontoctl

Was ist das?

Ein MCP-Server, CLI-Tool und eine TypeScript-Bibliothek für den Second-Hand-Marktplatz Vinted. Es gibt keine offizielle API — es erstellt ein Sitzungs-Cookie von der öffentlichen Katalogseite und ruft die private JSON-API auf, die die Vinted-Web-App intern verwendet.

  • 🤖 MCP-Server — verbinde dich mit Claude, Cursor oder einem beliebigen KI-Assistenten mit MCP-Unterstützung

  • 🖥️ CLI-Tool — Ergebnisse weiterleiten, neue Angebote beobachten, Preise direkt vom Terminal aus vergleichen

  • 📦 TypeScript-Bibliothek — importiere opSearch, opCompare usw. direkt in deinen Code


Installation

npm install -g @googlarz/vinted-client

Oder ohne Installation ausführen:

npx @googlarz/vinted-client search "levis 501"

CLI Schnellstart

# Search (JSON by default)
vinted search "levi's 501" --country fr

# Pretty table
vinted search "levi's 501" --country de --output table

# Filter by price, brand, condition
vinted search "adidas samba" \
  --price-min 20 --price-max 80 \
  --brand adidas \
  --condition new_with_tags,very_good \
  --output table

# Watch for new listings every 30s
vinted search "air jordan 1" --watch 30

# Walk all pages and collect up to 500 results
vinted search "vintage denim" --all --max-items 500

# Get a specific item (ID or URL)
vinted item 1234567
vinted item https://www.vinted.fr/items/1234567

# Seller profile + active listings
vinted seller 987654
vinted seller-items 987654 --output table

# Cross-country price comparison (6 countries by default)
vinted compare "north face jacket" --output table

# Browse category tree
vinted categories --query shoes --output table

# Look up brand IDs
vinted brands "stone island"

# What's trending right now
vinted trending --country fr --output table

Befehle

Befehl

Beschreibung

search <query>

Suche nach Angeboten mit voller Filterunterstützung

`item <id

url>`

Erhalte vollständige Artikeldetails

seller <id>

Verkäuferprofil

seller-items <id>

Artikel, die ein Verkäufer zum Verkauf anbietet

compare <query>

Preisvergleich über Länder hinweg

brands <query>

Suche Marken-IDs nach Namen

categories

Durchsuche den Kategorienbaum

trending

Neueste / trendige Angebote

debug

Überprüfe Sitzungs-Cookies (zur Fehlerbehebung)

Globale Flags

Flag

Beschreibung

`--output json

table`

Ausgabeformat (Standard: json)

--country <cc>

Ländercode (siehe unten)

--proxy <url>

HTTP/HTTPS-Proxy (auch: VINTED_PROXY_URL)

--no-cache

Deaktiviere den Antwort-Cache

Such-Flags

Flag

Beschreibung

--price-min / --price-max

Preisspanne

--brand <names>

Markennamen (automatisch in IDs aufgelöst)

--brand-ids <ids>

Kommagetrennte Marken-IDs

--category-id <n>

Kategorie-ID (vinted categories zum Durchsuchen)

--size-ids <ids>

Kommagetrennte Größen-IDs

--condition <list>

new_with_tags, new_without_tags, very_good, good, satisfactory

--sort <s>

relevance, price_low_to_high, price_high_to_low, newest_first

--date-from / --date-to

Datumsbereichsfilter (JJJJ-MM-TT)

--all

Seiten durchlaufen und alle Ergebnisse sammeln

--max-items <n>

Obergrenze für --all (Standard 1000)

--watch [interval]

Alle N Sekunden auf neue Angebote prüfen (Standard 60s)


Unterstützte Länder

fr de uk it es nl pl pt be at lt cz sk hu ro hr fi dk se


MCP Server

Binde Vinted in jeden MCP-kompatiblen KI-Assistenten ein (Claude, Cursor, etc.).

Einrichtung — Claude Desktop

Füge dies zu claude_desktop_config.json hinzu:

{
  "mcpServers": {
    "vinted": {
      "command": "npx",
      "args": ["-y", "@googlarz/vinted-client/mcp"]
    }
  }
}

Einrichtung — Claude Code

claude mcp add vinted -- npx -y @googlarz/vinted-client/mcp

MCP Tools

Tool

Beschreibung

search_items

Suche mit voller Filterunterstützung

get_item

Artikeldetails nach ID oder URL

get_seller

Verkäuferprofil

get_seller_items

Aktive Angebote eines Verkäufers

compare_prices

Preisvergleich über mehrere Länder

get_trending

Trendige Angebote

search_brands

Markensuche

get_categories

Kategorienbaum

Beispiel-Prompts nach der Verbindung:

"Suche mir Nike Air Max 95 unter 70 € in Deutschland, Größe 43, Zustand sehr gut"

"Vergleiche die Preise für eine North Face Pufferjacke in Frankreich, Deutschland und Italien"

"Beobachte Verkäufer #987654 und sag mir Bescheid, wenn sie etwas unter 30 € einstellen"


Bibliotheksnutzung

import { VintedClient, opSearch, opCompare, opSearchAll } from '@googlarz/vinted-client';

const client = new VintedClient();

// Basic search
const results = await opSearch(client, {
  query: 'levi\'s 501',
  country: 'de',
  priceMax: 50,
  condition: ['very_good', 'good'],
  sortBy: 'price_low_to_high',
});

console.log(results.items);

// Collect all pages concurrently (3-page prefetch window)
const all = await opSearchAll(client, {
  query: 'vintage band tee',
  country: 'uk',
  maxItems: 300,
});

// Multi-country price comparison
const report = await opCompare(client, {
  query: 'air jordan 1 retro',
  countries: ['fr', 'de', 'uk', 'it'],
});

Client-Optionen

const client = new VintedClient({
  proxyUrl: 'http://proxy:8080',   // or VINTED_PROXY_URL env var
  cacheTtlMs: 60_000,              // response cache TTL (0 = disable)
  rateLimitPerSec: 3,              // requests/sec per country
  rateLimitBurst: 6,               // burst capacity
  timeoutMs: 20_000,               // per-request timeout
});

Funktionsweise

Vinted hat keine öffentliche API. Diese Bibliothek:

  1. Startet eine Sitzung, indem sie vinted.{cc}/catalog aufruft und die Authentifizierungs-Cookies erfasst, die das Vinted-Frontend setzt.

  2. Ruft die private JSON-API (/api/v2/...) mit diesen Cookies auf und ahmt die Request-Header des Browsers nach.

  3. Startet bei 401 automatisch neu — Tokens laufen ab, die Bibliothek stellt die Verbindung im Hintergrund wieder her.

  4. Ratenbegrenzung pro Land mit einem Token-Bucket (konfigurierbarer Burst + Auffüllung), um 429-Fehler zu vermeiden.

  5. Cacht Antworten mit LRU+TTL — 60s für Suchergebnisse, 1h für statische Daten wie Kategorien.

  6. Greift auf HTML-Scraping zurück, wenn Artikelseiten durch DataDome blockiert werden (JSON-LD + Regex-Extraktion).

  7. Lädt 3 Seiten gleichzeitig vor in opSearchAll, um den Durchsatz innerhalb des Ratenlimits zu maximieren.


Proxy-Unterstützung

Wenn Vinted deine IP blockiert (häufig bei Cloud-VMs und CI), richte einen Proxy ein:

VINTED_PROXY_URL=http://user:pass@proxy:8080 vinted search "nike"
# or
vinted search "nike" --proxy http://user:pass@proxy:8080

Standard HTTPS_PROXY / HTTP_PROXY Umgebungsvariablen werden ebenfalls berücksichtigt.


Umgebungsvariablen

Variable

Beschreibung

VINTED_PROXY_URL

HTTP/HTTPS-Proxy-URL

VINTED_CACHE_TTL_MS

Cache-TTL in ms (Standard 60000)

VINTED_RATE_LIMIT_PER_SEC

Anfragen pro Sekunde pro Land (Standard 3)

VINTED_RATE_LIMIT_BURST

Token-Bucket-Burst-Größe (Standard 6)

VINTED_BROWSER

Auf 1 setzen, um einen Stealth-Browser für Artikeldetails zu verwenden


Anforderungen

  • Node.js ≥ 18

  • Optional: playwright + puppeteer-extra-plugin-stealth für den --browser / VINTED_BROWSER=1 Modus


Lizenz

MIT © googlarz


Available Tools

15 tools
compare_pricesA

Compare prices for a search query across multiple Vinted country sites simultaneously. Returns median, mean, stdDev, min, max, and sample count per country along with the local currency. Supports the same brand/size/color/condition filters as search_items for precise cross-country comparisons. Useful for finding the cheapest market or understanding cross-border price gaps.

Inputs: query (required), brand, limit, sizeIds, +5 more.

get_categoriesA

Fetch the full Vinted category tree for a country. Returns a flat list of all categories and subcategories with their numeric IDs, names, parent IDs, and item counts. Pass a categoryId to search_items or search_all_items to restrict results to a department (e.g. women's clothing, men's shoes, electronics). Results are cached for 1 hour.

Inputs: query, country.

get_colorsA

Fetch the complete list of Vinted color options available as search filters. Returns every color with its numeric ID, display name, hex color code, short code (e.g. "BLACK"), and sort order. Pass the returned IDs to search_items.colorIds or search_all_items.colorIds to restrict results to specific colors. Results are cached for 1 hour — color catalogues change rarely.

Inputs: country.

get_itemA

Fetch complete item details by Vinted item ID (with country) or by a direct Vinted item URL. Returns title, price, currency, brand, size, condition, full description, all photo URLs, creation date, item URL, favourite count, and seller username/ID. Automatically falls back to HTML scraping when the JSON API is unavailable.

Inputs: url, itemId, country.

get_new_itemsA

Fetch the most recently listed items for a search query (newest first). To monitor a search, call it repeatedly and pass the previous response's latestId as afterId — only items listed since then are returned (truncated: true means more than perPage arrived; raise perPage or poll more often). Supports the same filters as search_items (brand, category, size, color, price, condition).

Inputs: query (required), size, brand, color, +10 more.

get_sellerA

Fetch a seller's public profile by their numeric user ID. Returns username, active listing count, feedback reputation score (0–1 float), total feedback count, country code, and profile URL. Use get_seller_feedback to read review texts and star ratings, and get_seller_items to browse their listings.

Inputs: sellerId (required), country.

get_seller_feedbackA

Fetch paginated buyer and seller feedback reviews for a Vinted user. Each entry includes the review text, star rating (1–5), feedback type (1=negative, 2=neutral, 3=positive), reviewer username, timestamp, and the associated item ID. Use this to assess seller trustworthiness and reliability before making a purchase.

Inputs: sellerId (required), page, limit, country.

get_seller_itemsA

List all items currently for sale by a specific seller, paginated. Returns the same fields as search_items (title, price, brand, size, condition, photo URL, item URL). Useful for browsing a seller's full catalogue after finding them via search_items or get_seller.

Inputs: sellerId (required), page, limit, country.

get_size_groupsA

Fetch all Vinted size groups with their constituent size IDs and labels. Returns a list of size groups (e.g. "Women's clothing", "Men's shoes", "Kids 2–8 yrs") — each containing the group ID, caption, description, and an array of sizes with numeric IDs and display titles (e.g. "XS", "42", "12 UK"). Pass individual size IDs from the sizes array to search_items.sizeIds or search_all_items.sizeIds to filter listings to an exact size. Results are cached for 1 hour.

Inputs: country.

resolve_color_idsA

Resolve color names (e.g. "black", "white", "navy") to numeric Vinted color IDs for use in search_items.colorIds and search_all_items.colorIds. Matching is case-insensitive and also works with Vinted's internal color codes (e.g. "BLACK", "NAVY_BLUE"). Alternatively, pass color names directly to search_items.color[] and they will be resolved automatically.

Inputs: colors (required), country.

resolve_size_idsA

Resolve human-readable size labels (e.g. "M", "42", "XL", "12 UK") to numeric Vinted size IDs suitable for use in search_items.sizeIds and search_all_items.sizeIds. Returns matched IDs with the group they belong to (e.g. "Women's clothing"), and a list of any labels that could not be resolved. Because "M" or "42" appear in multiple size groups (women's, men's, kids'), multiple IDs may be returned per label — pass them all to sizeIds to cast a wide net, or filter by group name if you need a specific category.

Inputs: sizes (required), country.

search_all_itemsA

Search Vinted listings and automatically paginate through all results, returning up to maxItems items in a single call. Use this instead of search_items when you need comprehensive results — e.g. "find all Nike shoes under €30" or "list every item in size M from this brand". Pages are fetched concurrently for speed. Returns the same item fields as search_items.

Inputs: query (required), size, brand, color, +11 more.

search_brandsA

Search Vinted's brand catalogue by keyword. Returns matching brands with their numeric IDs, slugs, total item counts, and favourite counts. Pass the returned IDs to search_items.brandIds, or use search_items.brand[] to pass names and have them resolved automatically.

Inputs: query (required), limit, country.

search_itemsA

Search Vinted second-hand listings with rich filters across 19 country sites. Returns a paginated list of items — each with title, price, currency, brand, size, condition, photo URL, item URL, favourite count, and seller info. Use get_categories to discover valid categoryId values and search_brands to resolve brand names to IDs. For comprehensive multi-page results use search_all_items instead.

Inputs: query (required), page, size, brand, +11 more.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A4.2/5.0

    Scored across 15 tools

    Disambiguation4/5

    Most tools target a distinct resource or action and the descriptions actively cross-reference each other (e.g. search_items points to search_all_items, resolve_color_ids points to get_colors). The main confusable pair is get_trending vs get_new_items, since both return recency-ordered items, with only country-scoping vs query+monitoring to tell them apart; the resolve_* vs get_* catalog pairs are related but functionally distinct.

    Naming Consistency5/5

    Every tool follows a clean snake_case verb_noun pattern (search_items, get_item, compare_prices, resolve_size_ids, get_seller_feedback). Verbs are used consistently and predictably across the whole set.

    Tool Count5/5

    15 tools is well scoped for a marketplace search/browse server, with each tool covering a distinct capability (search, detail, price comparison, taxonomy lookups, seller data). Nothing feels padded or redundant.

    Completeness4/5

    Coverage is strong for a read-only browsing server: search with rich filters, item details, cross-country price comparison, trending/new items, taxonomy lookups (categories, colors, sizes, brands), and seller profile/feedback/items. Minor gaps exist since it exposes favourite counts but offers no watchlist or seller-messaging actions.

    Maintenance

    ActivityMaintained
    ResponsivenessOutdated data

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      D
      maintenance
      This MCP scraps vinted for product info. Disclaimer: This script is designed for educational purposes only. It is intended to demonstrate web scraping techniques and should not be used for any commercial or personal gain. Please note that using this software may violate the terms of service of Vint
      149
      GPL 3.0
    • A
      license
      Not graded
      quality
      C
      maintenance
      An MCP server for Vinted search and analysis that provides tools to search listings, fetch item details, inspect seller profiles, compare prices across countries, and surface trending items.
      163 npm
      36
      AGPL 3.0