Skip to main content
Glama
oxylabs

Oxylabs MCP Server

Official

pypi package Licence Coverage badge


📖 Übersicht

Der Oxylabs-MCP-Server stellt eine Brücke zwischen KI-Modellen und dem Web dar. Er ermöglicht es ihnen, beliebige URLs zu scrapen, JavaScript-lastige Seiten zu rendern, Inhalte für die KI-Nutzung zu extrahieren und zu formatieren, CAPTCHAs zu verwalten und auf geobeschränkte Webdaten aus über 195 Ländern zuzugreifen.

Er basiert auf dem Model Context Protocol (MCP), dem offenen Standard für die Verbindung von KI-Assistenten mit externen Tools und Daten.

Related MCP server: FreeCrawl MCP Server

🛠️ MCP-Tools

Oxylabs MCP stellt zwei Toolsets bereit, die zusammen oder unabhängig voneinander verwendet werden können:

Oxylabs-Web-Scraper-API-Tools

  1. universal_scraper: scrapet beliebige URLs, mit optionalem JavaScript-Rendering, Geo-Targeting und Markdown/HTML/Links-Ausgabe;

  2. google_search_scraper: extrahiert Ergebnisse aus der Google-Suche, mit optionalem Parsing in strukturiertes JSON;

  3. amazon_search_scraper: scrapet Amazon-Suchergebnisseiten, mit optionalem Parsing in strukturiertes JSON;

  4. amazon_product_scraper: extrahiert Daten aus einzelnen Amazon-Produktseiten.

Oxylabs-AI-Studio-Tools

  1. ai_scraper: scrapet Inhalte von beliebigen URLs mit KI-gestützter Extraktion, im JSON-, CSV-, Markdown- oder TOON-Format;

  2. ai_crawler: durchsucht eine Website ausgehend von einer Start-URL basierend auf einem Prompt und sammelt Daten über mehrere Seiten;

  3. ai_browser_agent: steuert einen echten Browser basierend auf einem Prompt – navigiert, klickt, füllt Formulare aus – und gibt das Ergebnis zurück;

  4. ai_search: durchsucht das Web und gibt optional den Markdown-Inhalt jedes Ergebnisses zurück;

  5. ai_map: kartiert die URLs einer Website, gefiltert nach Schlüsselwörtern oder einem Prompt;

  6. generate_schema: generiert ein JSON-Schema im OpenAPI-Format für die strukturierte Extraktion mit den oben genannten KI-Tools.

✅ Voraussetzungen

Bevor Sie beginnen, stellen Sie sicher, dass Sie mindestens eines der Folgenden besitzen:

  • Oxylabs-Web-Scraper-API-Konto: Erhalten Sie Ihren Benutzernamen und Ihr Passwort von Oxylabs (kostenlose 1-Wochen-Testversion verfügbar);

  • Oxylabs-AI-Studio-API-Schlüssel: Erhalten Sie Ihren API-Schlüssel von Oxylabs AI Studio (1000 Credits kostenlos).

Um den Server lokal auszuführen (Option 2 unten), benötigen Sie außerdem den Paketmanager uv:

# macOS and Linux
curl -LsSf https://astral.sh/uv/install.sh | sh
# Windows
powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

📦 Konfiguration

Es gibt zwei Möglichkeiten, den Server zu nutzen: Verbinden Sie sich mit der gehosteten Instanz (keine Installation) oder führen Sie ihn lokal mit Anmeldedaten in Umgebungsvariablen aus.

Option 1: Gehosteter Server (keine Installation)

Oxylabs betreibt einen gehosteten MCP-Server unter:

https://mcp.oxylabs.io/mcp

Anmeldedaten werden über Request-Header übergeben:

Anmeldedaten

Header

Web Scraper API

Authorization: Basic <base64(username:password)>

Web Scraper API (alternativ)

X-Oxylabs-Username und X-Oxylabs-Password

AI Studio

X-Oxylabs-AI-Studio-Api-Key

Einrichtung mit Claude Code:

claude mcp add --transport http oxylabs https://mcp.oxylabs.io/mcp \
  --header "Authorization: Basic $(echo -n 'YOUR_USERNAME:YOUR_PASSWORD' | base64)" \
  --header "X-Oxylabs-AI-Studio-Api-Key: YOUR_API_KEY"

Einrichtung mit Cursor oder einem beliebigen Client, der Remote-MCP-Server mit benutzerdefinierten Headern unterstützt:

{
  "mcpServers": {
    "oxylabs": {
      "url": "https://mcp.oxylabs.io/mcp",
      "headers": {
        "Authorization": "Basic <base64 of username:password>",
        "X-Oxylabs-AI-Studio-Api-Key": "YOUR_API_KEY"
      }
    }
  }
}

Der Server ist auch auf Smithery gelistet.

Hinweis: Clients, die für Remote-Server nur OAuth unterstützen (z. B. das Hinzufügen eines benutzerdefinierten Connectors in der claude.ai-Weboberfläche), können noch keine Header übergeben – die OAuth-Anmeldung steht auf unserer Roadmap. Verwenden Sie in der Zwischenzeit die lokale Einrichtung unten mit diesen Clients.

Option 2: Lokal ausführen

Umgebungsvariablen

Der Oxylabs-MCP-Server unterstützt die folgenden Umgebungsvariablen:

Name

Beschreibung

Standard

OXYLABS_USERNAME

Ihr Oxylabs-Web-Scraper-API-Benutzername

OXYLABS_PASSWORD

Ihr Oxylabs-Web-Scraper-API-Passwort

OXYLABS_AI_STUDIO_API_KEY

Ihr Oxylabs-AI-Studio-API-Schlüssel

LOG_LEVEL

Protokollierungsgrad für die an den Client zurückgegebenen Protokolle

INFO

Basierend auf den bereitgestellten Anmeldedaten stellt der Server automatisch die entsprechenden Tools bereit:

  • Wenn nur OXYLABS_USERNAME und OXYLABS_PASSWORD angegeben werden, stellt der Server die Web-Scraper-API-Tools bereit;

  • Wenn nur OXYLABS_AI_STUDIO_API_KEY angegeben wird, stellt der Server die AI-Studio-Tools bereit;

  • Wenn alle drei angegeben werden, stellt der Server alle Tools bereit.

❗ Wichtig: Setzen Sie nur die Umgebungsvariablen, für die Sie echte Anmeldedaten besitzen. Platzhalterwerte führen dazu, dass Tools bereitgestellt werden, die nicht funktionieren.

Mit uvx konfigurieren

Installiert das Paket von PyPI und führt es automatisch aus:

{
  "mcpServers": {
    "oxylabs": {
      "command": "uvx",
      "args": ["oxylabs-mcp"],
      "env": {
        "OXYLABS_USERNAME": "YOUR_USERNAME",
        "OXYLABS_PASSWORD": "YOUR_PASSWORD",
        "OXYLABS_AI_STUDIO_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

Mit einem lokalen Checkout konfigurieren

Nützlich für die Entwicklung – führt den Server aus einem lokalen Klon dieses Repositorys aus:

{
  "mcpServers": {
    "oxylabs": {
      "command": "uv",
      "args": [
        "--directory",
        "/<absolute-path-to-folder>/oxylabs-mcp",
        "run",
        "oxylabs-mcp"
      ],
      "env": {
        "OXYLABS_USERNAME": "YOUR_USERNAME",
        "OXYLABS_PASSWORD": "YOUR_PASSWORD",
        "OXYLABS_AI_STUDIO_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

Als Remote-HTTP-Server ausführen (Self-Hosting)

Der Server unterstützt auch den MCP-Streamable-HTTP-Transport. Starten Sie ihn mit:

MCP_TRANSPORT=streamable-http MCP_HOST=0.0.0.0 MCP_PORT=8000 uvx oxylabs-mcp

Beim HTTP-Transport werden Anmeldedaten pro Request übergeben, statt über Umgebungsvariablen:

Anmeldedaten

So übergeben Sie sie

Web Scraper API

Authorization: Basic <base64(username:password)> (Standard-HTTP-Basic-Auth)

Web Scraper API (alternativ)

X-Oxylabs-Username- und X-Oxylabs-Password-Header

AI Studio

X-Oxylabs-AI-Studio-Api-Key-Header

Beispiel-Client-Konfiguration:

{
  "mcpServers": {
    "oxylabs": {
      "url": "https://your-host:8000/mcp",
      "headers": {
        "Authorization": "Basic <base64 of username:password>",
        "X-Oxylabs-AI-Studio-Api-Key": "YOUR_API_KEY"
      }
    }
  }
}

Alle Tools werden unabhängig von den bereitgestellten Anmeldedaten immer aufgelistet; der Aufruf eines Tools ohne die benötigten Anmeldedaten gibt eine Fehlermeldung zurück, die genau erklärt, was zu konfigurieren ist.

Einrichtung mit Claude Desktop

Navigieren Sie zu Claude → Einstellungen → Entwickler → Konfiguration bearbeiten und fügen Sie eine der obigen Konfigurationen zur Datei claude_desktop_config.json hinzu.

Einrichtung mit Cursor AI

Navigieren Sie zu Cursor → Einstellungen → Cursor-Einstellungen → MCP. Klicken Sie auf Neuen globalen MCP-Server hinzufügen und fügen Sie eine der obigen Konfigurationen hinzu.

📝 Protokollierung

Der Server stellt zusätzliche Informationen über die Tool-Aufrufe in notification/message-Ereignissen bereit:

{
  "method": "notifications/message",
  "params": {
    "level": "info",
    "data": "Create job with params: {\"url\": \"https://ip.oxylabs.io\"}"
  }
}
{
  "method": "notifications/message",
  "params": {
    "level": "info",
    "data": "Job info: job_id=7333113830223918081 job_status=done"
  }
}
{
  "method": "notifications/message",
  "params": {
    "level": "error",
    "data": "Error: request to Oxylabs API failed"
  }
}

✨ Wichtige Funktionen

  • Daten aus beliebigen URLs extrahieren, einschließlich komplexer Single-Page-Anwendungen

  • Dynamische Websites mit Headless-Browser-Unterstützung vollständig rendern

  • Zwischen vollständigem JavaScript-Rendering, nur HTML oder keiner Rendering-Option wählen

  • Mobile- und Desktop-Viewports für realistisches Rendering emulieren

  • HTML automatisch bereinigen und für bessere Lesbarkeit in Markdown konvertieren

  • Automatisierte Parser für beliebte Ziele wie Google, Amazon und mehr verwenden

  • Anspruchsvolle automatisierte Request-Management-Systeme mit hoher Erfolgsquote umgehen

  • Auch die komplexesten Websites zuverlässig scrapen

  • Automatisch rotierende IPs aus einem Proxy-Pool mit Abdeckung von über 195 Ländern erhalten

  • Bei Bedarf Rendering- und Parsing-Optionen festlegen

  • Daten direkt in KI-Modelle oder Analysetools einspeisen

  • Funktioniert unter macOS, Windows und Linux

  • Umfassende Fehlerbehandlung und -meldung

  • Intelligentes Rate Limiting und Request-Verwaltung

Warum Oxylabs MCP?  🕸️ ➜ 📦 ➜ 🤖

Stellen Sie sich vor, Sie sagen Ihrem LLM „Fasse die neueste Hacker-News-Diskussion über GPT‑5 zusammen" – und es antwortet einfach. Der Oxylabs-MCP-Server macht das möglich, indem er die langweiligen Teile für Sie erledigt:

Was Oxylabs MCP tut

Warum es für Sie wichtig ist

Verwaltet automatisierte Request-Wände mit dem globalen Oxylabs-Proxy-Netzwerk

Ermöglicht Website-Zugriff und Anonymität

Rendert JavaScript in Headless-Chrome

Single-Page-Apps, erledigt

Bereinigt HTML → Markdown

Direkt in Vektor-Datenbanken oder Prompts einfügen

Optionale strukturierte Parser (Google, Amazon usw.)

Zugriff auf beliebte Ziele mit einer Zeile


🛡️ Lizenz

Verteilt unter der MIT-Lizenz – siehe LICENSE für Details.


Über Oxylabs

Oxylabs wurde 2015 gegründet und ist eine marktführende Plattform für die Erfassung von Web-Intelligence, die von den höchsten Geschäfts-, Ethik- und Compliance-Standards angetrieben wird und Unternehmen weltweit datengetriebene Erkenntnisse ermöglicht.

image

mcp-name: io.oxylabs/oxylabs-mcp

Available Tools

10 tools
ai_browser_agentC
Read-only
Inspect

Run the browser agent and return the data in the specified format.

This tool is useful if you need navigate around the website and do some actions. It allows navigating to any url, clicking on links, filling forms, scrolling, etc. Finally it returns the data in the specified format. Schema is required only if output_format is json, csv or toon. 'task_prompt' describes what browser agent should achieve

ParametersJSON Schema
NameRequiredDescriptionDefault
urlYesThe URL to start the browser agent navigation from.
schemaNoThe schema to use for the scrape. Only required if output_format is json, csv or toon.
task_promptYesWhat browser agent should do.
geo_locationNoTwo letter ISO country code to use for the browser proxy.
output_formatNoThe output format. Markdown returns full text of the page including links. Toon(Token-Oriented Object Notation) returns data in Toon format, which is optimized for AI agents. If json, csv or toon, the schema is required.markdown

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

C2.9/5.0
Behavior1/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

The description states the agent can click links, fill forms, and perform actions, which implies potentially mutating state or submitting data. This contradicts the annotations readOnlyHint set to true. Because the description directly conflicts with the annotation and also provides no safety/auth/side-effect context, this dimension scores minimal.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness3/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is not excessively long, but it repeats the core idea: the first sentence and the 'Finally it returns...' sentence both say the tool returns data in a specified format. Some sentences are redundant rather than adding new operational guidance.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

The output schema and detailed input-schema provide substantial structure, so this is not a sparse definition. However, the description misses important behavioral context around a browser automation tool, such as side effects from form submission, authentication state, and session behavior; this is made worse by the annotation contradiction.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The schema already covers all parameters with descriptions at 100% coverage, so the baseline is 3. The description repeats the conditional schema requirement for json/csv/toon and explains task_prompt, but it adds no new information beyond what the schema provides.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly says 'Run the browser agent and return the data' and then lists concrete actions like clicking, filling forms, scrolling, and navigating to URLs. This gives a specific verb/resource and conveys an interactive browser tool, though it does not explicitly name or contrast sibling scraper/crawler tools.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description says the tool is 'useful if you need navigate around the website and do some actions,' which implies an interactive task. However, it provides no explicit guidance on when not to use it or which sibling tool (e.g., ai_scraper, ai_crawler) should be used for static extraction.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

ai_crawlerC
Read-only
Inspect

Tool useful for crawling a website from starting url and returning data in a specified format.

Schema is required only if output_format is json, csv or toon. 'render_javascript' is used to render javascript heavy websites. 'return_sources_limit' is used to limit the number of sources to return, for example if you expect results from single source, you can set it to 1.

ParametersJSON Schema
NameRequiredDescriptionDefault
urlYesThe URL from which crawling will be started.
schemaNoThe JSON schema to use for structured data extraction from the crawled pages. Only required if output_format is json, csv or toon.
user_promptYesWhat information user wants to extract from the domain.
geo_locationNoTwo letter ISO country code to use for the crawl proxy.
output_formatNoThe format of the output. If json, csv or toon, the schema is required. Markdown returns full text of the page. CSV returns data in CSV format. Toon(Token-Oriented Object Notation) returns data in Toon format, which is optimized for AI agents.markdown
render_javascriptNoWhether to render the HTML of the page using javascript. Much slower, therefore use it only for websites that require javascript to render the page. Unless user asks to use it, first try to crawl the page without it. If results are unsatisfactory, try to use it.
return_sources_limitNoThe maximum number of sources to return.

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

C2.9/5.0
Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations only provide readOnlyHint=true, which the description respects. The description does not add further behavioral details (e.g., no side effects, rate limits, or data retention), but it does not contradict the annotation either. Given the read-only nature is already indicated, the description adds little beyond that.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness2/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is verbose and redundant, repeating parameter details that are already in the schema. For example, the URL and render_javascript explanations are duplicated verbatim. This wastes tokens and reduces clarity.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

An output schema is present but the description does not explain the structure or any exceptional behaviors. It briefly mentions returning data in a specified format, but does not elaborate on how the crawl is scoped or what happens with large sites. Given the completeness of the input schema and presence of output schema, the description is adequate but not thorough.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the tool description adds no new parameter information. The prose repeats the schema definitions without clarifying edge cases or relationships, so it meets the baseline but provides no added value.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool crawls a website starting from a URL and returns data in a specified format. However, it does not differentiate from sibling tools like ai_scraper or universal_scraper, which might also crawl pages.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description does not explicitly state when to use this tool over alternatives. It implies output format flexibility but lacks guidance on scenarios favoring ai_crawler over other scrapers or search tools.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

ai_mapD
Read-only
Inspect

Tool useful for mapping website's URLs.

ParametersJSON Schema
NameRequiredDescriptionDefault
urlYesThe URL from which URLs mapping will be started.
limitNoThe maximum number of URLs to return.
user_promptNoWhat kind of URLs user wants to find. Can be used together with 'search_keywords'.
geo_locationNoTwo letter ISO country code to use for the mapping proxy.
max_crawl_depthNoThe maximum depth of the crawl.
search_keywordsNoThe keywords to use for URLs paths filtering. Keywords are matched as OR condition. Meaning, one keyword is enough to match the url path.
allow_subdomainsNoWhether to map subdomains URLs as well.
render_javascriptNoWhether to render the HTML of the page using javascript. Much slower, therefore use it only for websites that require javascript to render the page. Unless user asks to use it, first try to crawl the page without it. If results are unsatisfactory, try to use it.
allow_external_domainsNoWhether to include external domains URLs.

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

D1.9/5.0
Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

The annotations declare readOnlyHint: true, indicating a safe read operation, but the description fails to add behavioral context. It doesn't disclose that the tool performs crawling, respects depth limits, or requires JavaScript rendering for some sites. No mention of performance characteristics, rate limits, or edge cases beyond what the schema's parameter descriptions already provide.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness2/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The single-sentence description is concise but under-specifies the tool's behavior. It's not powerful or informative enough to earn credit for conciseness; rather, it reads as an under-developed placeholder. A good description would front-load the tool's purpose in a way that adds value, but this wastes the opportunity.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness1/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a tool with 9 parameters, a required URL, and complex behaviors like crawl depth, subdomain handling, and JavaScript rendering, a one-sentence description is grossly inadequate. Even though an output schema exists, the description fails to convey the tool's inputs' intent or the meaning of its output. The behavioral nuances (e.g., proxy usage, OR-matching for keywords) are left entirely to the schema, making this incomplete.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so all 9 parameters are documented in the input schema itself. The description adds no parameter semantics; it merely repeats the tool name. Per rubric, with high schema coverage, the baseline is 3, which is appropriate here since the description doesn't need to compensate but also doesn't add value.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose2/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description 'Tool useful for mapping website's URLs' uses the vague verb 'mapping' and a possessive phrasing that doesn't define the action clearly. While it names the resource (website URLs), it fails to articulate the core function of discovering or crawling links, leaving the tool's true purpose ambiguous. Sibling tools like 'ai_crawler' and 'ai_scraper' further blur the line, making this description insufficiently specific.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines1/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides no guidance on when to use this tool versus any of the nine sibling tools. There is no mention of when ai_map is preferred over ai_crawler or ai_browser_agent, nor any exclusions or prerequisites. Users are left to guess which tool fits their use case.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

ai_scraperA
Read-only
Inspect

Scrape the contents of the web page and return the data in the specified format.

Schema is required only if output_format is json or csv. 'render_javascript' is used to render javascript heavy websites.

ParametersJSON Schema
NameRequiredDescriptionDefault
urlYesThe URL to scrape
schemaNoThe JSON schema to use for structured data extraction from the scraped page. Only required if output_format is json, csv or toon.
geo_locationNoTwo letter ISO country code to use for the scrape proxy.
output_formatNoThe format of the output. If json, csv or toon, the schema is required. Markdown returns full text of the page. CSV returns data in CSV format, tabular like data. Toon(Token-Oriented Object Notation) returns data in Toon format, which is optimized for AI agents.markdown
render_javascriptNoWhether to render the HTML of the page using javascript. Much slower, therefore use it only for websites that require javascript to render the page.Unless user asks to use it, first try to scrape the page without it. If results are unsatisfactory, try to use it.

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

A3.6/5.0
Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations declare readOnlyHint=true, and the description aligns with that (scraping is read-only). The description adds useful behavioral context about render_javascript being slower and the recommendation to try without it first. However, it doesn't disclose potential rate limits, auth requirements, or what happens on failure, which would be valuable for a scraping tool.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is concise (two sentences) and front-loaded with the core purpose. It avoids redundancy with the schema. However, it could be slightly more structured by separating the conditional requirements more clearly, but overall it's efficient.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool has an output schema and 100% parameter coverage, the description is fairly complete. It covers the key conditional logic (schema requirement, render_javascript usage) and the tool's scope. It doesn't explain return values, but the output schema handles that. Minor gaps: no mention of error handling or edge cases, but acceptable for a scraping tool.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the schema already documents all parameters well. The description adds minimal extra meaning beyond what the schema provides, but it does clarify the conditional requirement for schema and the performance trade-off of render_javascript. This is a baseline 3 since the schema does the heavy lifting.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool scrapes web page contents and returns data in a specified format. It distinguishes itself from siblings like ai_crawler (which likely crawls multiple pages) and google_search_scraper (which targets search results) by focusing on a single page scrape with format options.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides clear guidance on when schema is required (for json/csv/toon formats) and when to use render_javascript (for JS-heavy sites, with a recommendation to try without it first). It doesn't explicitly mention alternatives among siblings, but the usage context is well-defined.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

amazon_product_scraperA
Read-only
Inspect

Scrape Amazon products.

Supports content parsing, different user agent types, domain, geolocation, locale parameters and different output formats. Supports Amazon specific parameters such as currency and getting more accurate pricing data with auto select variant.

ParametersJSON Schema
NameRequiredDescriptionDefault
parseNoShould result be parsed. If the result is not parsed, the output_format parameter is applied.
queryYesKeyword to search for.
domainNo Domain localization for Google. Use country top level domains. For example: - 'co.uk' for United Kingdom - 'us' for United States - 'fr' for France
localeNo Set 'Accept-Language' header value which changes your Google search page web interface language. Examples: - 'en-US' for English, United States - 'de-AT' for German, Austria - 'fr-FR' for French, France
renderNo Whether a headless browser should be used to render the page. For example: - 'html' when browser is required to render the page.
currencyNoCurrency that will be used to display the prices.
geo_locationNo The geographical location that the result should be adapted for. Use ISO-3166 country codes. Examples: - 'California, United States' - 'Mexico' - 'US' for United States - 'DE' for Germany - 'FR' for France
output_formatNo The format of the output. Works only when parse parameter is false. - links - Most efficient when the goal is navigation or finding specific URLs. Use this first when you need to locate a specific page within a website. - md - Best for extracting and reading visible content once you've found the right page. Use this to get structured content that's easy to read and process. - html - Should be used sparingly only when you need the raw HTML structure, JavaScript code, or styling information.
user_agent_typeNoDevice type and browser that will be used to determine User-Agent header value.
autoselect_variantNoTo get accurate pricing/buybox data, set this parameter to true.

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

A3.5/5.0
Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

The readOnlyHint annotation already signals a safe read operation. The description adds some functional context (content parsing, output formats, user agents) but does not disclose potential side effects, rate limits, or return behavior beyond what the schema implies. The annotation is not contradicted, and the added detail provides marginal value.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is two sentences with the purpose front-loaded. The second sentence compactly enumerates capabilities without redundancy. Every word contributes to the overall understanding, and there is no filler or unnecessary detail.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's complexity (10 parameters, 1 required) and rich schema/output schema, the description gives a reasonable high-level overview. However, it does not mention limitations, pagination, or differentiate from similar tools like amazon_search_scraper. The output schema and annotations fill in some gaps, making it minimally complete but not richly contextual.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema has 100% description coverage for all 10 parameters, so the schema carries the full semantic burden. The description groups parameters into categories and highlights Amazon-specific ones (currency, autoselect_variant), adding conceptual organization but no new factual information beyond the schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool scrapes Amazon products, providing a specific verb and resource. It does not explicitly differentiate from sibling tools like amazon_search_scraper or universal_scraper, so it stops short of a 5.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description implies the tool is for scraping Amazon products but gives no explicit guidance on when to use it vs. alternatives. It lacks when-to-use/when-not-to-use criteria or exclusion notes, leaving the agent to infer based on the name and general purpose.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

amazon_search_scraperB
Read-only
Inspect

Scrape Amazon search results.

Supports content parsing, different user agent types, pagination, domain, geolocation, locale parameters and different output formats. Supports Amazon specific parameters such as category id, merchant id, currency.

ParametersJSON Schema
NameRequiredDescriptionDefault
pagesNoNumber of pages to retrieve.
parseNoShould result be parsed. If the result is not parsed, the output_format parameter is applied.
queryYesKeyword to search for.
domainNo Domain localization for Google. Use country top level domains. For example: - 'co.uk' for United Kingdom - 'us' for United States - 'fr' for France
localeNo Set 'Accept-Language' header value which changes your Google search page web interface language. Examples: - 'en-US' for English, United States - 'de-AT' for German, Austria - 'fr-FR' for French, France
renderNo Whether a headless browser should be used to render the page. For example: - 'html' when browser is required to render the page.
currencyNoCurrency that will be used to display the prices.
start_pageNoStarting page number.
category_idNoSearch for items in a particular browse node (product category).
merchant_idNoSearch for items sold by a particular seller.
geo_locationNo The geographical location that the result should be adapted for. Use ISO-3166 country codes. Examples: - 'California, United States' - 'Mexico' - 'US' for United States - 'DE' for Germany - 'FR' for France
output_formatNo The format of the output. Works only when parse parameter is false. - links - Most efficient when the goal is navigation or finding specific URLs. Use this first when you need to locate a specific page within a website. - md - Best for extracting and reading visible content once you've found the right page. Use this to get structured content that's easy to read and process. - html - Should be used sparingly only when you need the raw HTML structure, JavaScript code, or styling information.
user_agent_typeNoDevice type and browser that will be used to determine User-Agent header value.

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

B3.4/5.0
Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

The description only lists supported features (parsing, pagination, user agents, etc.) and does not disclose behavioral traits such as response format, rate limits, or edge cases. The readOnlyHint annotation already indicates a safe read operation, but the description adds little beyond what the schema and annotation provide.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is concise, with a clear opening statement followed by a feature list in two sentences. While every sentence provides relevant information, the list format is somewhat generic and could be better structured by separating capabilities into categories.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's complexity (13 parameters) and the presence of a full output schema, the description offers an adequate high-level overview. However, it omits practical context like when to use parse=false or how pagination behaves, relying on the detailed schema descriptions to cover specifics.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the baseline is 3. The description redundantly mentions parameter groups already documented in the schema (e.g., pagination, user agent types, currency) without adding nuanced meaning or context beyond what the schema descriptions offer.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description begins with a specific verb and resource: 'Scrape Amazon search results.' This clearly distinguishes it from siblings like google_search_scraper and amazon_product_scraper, which target different resources. It further lists Amazon-specific parameters, reinforcing its focused purpose.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description implies usage for Amazon search result scraping but provides no explicit guidance on when to choose this tool over alternatives. It does not mention exclusions or recommend siblings for related tasks, leaving the decision to inference from the tool name and capability list.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

generate_schemaC
Read-only
Inspect

Generate a json schema in openapi format.

ParametersJSON Schema
NameRequiredDescriptionDefault
app_nameYes
user_promptYes

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

C2.2/5.0
Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

The description does not contradict the readOnly annotation, but it adds no context about side effects, limitations, or special behaviors. With the annotation present, the bar is lower, but the description still offers minimal insight beyond the tool's name.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness3/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single, concise sentence with no fluff, but it is too brief to be informative. It is appropriately sized in terms of length, but the lack of content reduces its effectiveness.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness1/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool has two parameters and an output schema, the description is severely incomplete. It does not explain expected inputs, outputs, or any relevant context, making it insufficient for a user to understand the tool's full capabilities.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters1/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The description gives no explanation of the parameters user_prompt and app_name. Schema coverage is 0%, and the description fails to compensate with any param-level detail, leaving the user to guess their meaning.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool generates a JSON schema in OpenAPI format, which is a specific action and outcome. It differentiates from sibling tools focused on search and scraping, but could be more specific about the schema's intended use.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines1/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides no guidance on when to use this tool versus alternatives, nor any conditions or prerequisites. It lacks explicit when-to-use or when-not-to-use information.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

google_search_scraperA
Read-only
Inspect

Scrape Google Search results.

Supports content parsing, different user agent types, pagination, domain, geolocation, locale parameters and different output formats.

ParametersJSON Schema
NameRequiredDescriptionDefault
limitNoNumber of results to retrieve in each page.
pagesNoNumber of pages to retrieve.
parseNoShould result be parsed. If the result is not parsed, the output_format parameter is applied.
queryYesURL-encoded keyword to search for.
domainNo Domain localization for Google. Use country top level domains. For example: - 'co.uk' for United Kingdom - 'us' for United States - 'fr' for France
localeNo Set 'Accept-Language' header value which changes your Google search page web interface language. Examples: - 'en-US' for English, United States - 'de-AT' for German, Austria - 'fr-FR' for French, France
renderNo Whether a headless browser should be used to render the page. For example: - 'html' when browser is required to render the page.
ad_modeNoIf true will use the Google Ads source optimized for the paid ads.
start_pageNoStarting page number.
geo_locationNo The geographical location that the result should be adapted for. Use ISO-3166 country codes. Examples: - 'California, United States' - 'Mexico' - 'US' for United States - 'DE' for Germany - 'FR' for France
output_formatNo The format of the output. Works only when parse parameter is false. - links - Most efficient when the goal is navigation or finding specific URLs. Use this first when you need to locate a specific page within a website. - md - Best for extracting and reading visible content once you've found the right page. Use this to get structured content that's easy to read and process. - html - Should be used sparingly only when you need the raw HTML structure, JavaScript code, or styling information.
user_agent_typeNoDevice type and browser that will be used to determine User-Agent header value.

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

A3.7/5.0
Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations declare readOnlyHint=true, and the description 'Scrape' is consistent with a read operation. However, the description adds little beyond the annotations and the parameter schema; it does not mention rate limits, pagination behavior, rendering implications, or antiscraping nuances.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is only two sentences. The first sentence is a clear, front-loaded purpose statement; the second is a compact capability list. There is no filler or redundancy.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

The schema and output schema are rich, but the tool description omits several notable parameters like render and ad_mode, and does not explain the parse/output_format relationship. Given the tool's complexity, the description alone provides only high-level context, leaving these gaps.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the schema fully documents all 12 parameters. The tool description only lists categories like 'pagination' and 'geolocation' without adding new meaning; therefore, a baseline 3 is appropriate.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description starts with a specific verb and resource: 'Scrape Google Search results.' This clearly distinguishes the tool from siblings like amazon_search_scraper or ai_search by naming Google Search as the target. The second sentence enumerates key capabilities (parsing, user agents, pagination, etc.), further clarifying scope.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description implies usage for scraping Google Search results but provides no explicit guidance on when to prefer this over ai_search, universal_scraper, or other siblings. It lists supported features but does not state conditions, exclusions, or alternative choices.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

universal_scraperB
Read-only
Inspect

Get a content of any webpage.

Supports browser rendering, parsing of certain webpages and different output formats.

ParametersJSON Schema
NameRequiredDescriptionDefault
urlYesWebsite url to scrape.
renderNo Whether a headless browser should be used to render the page. For example: - 'html' when browser is required to render the page.
geo_locationNo The geographical location that the result should be adapted for. Use ISO-3166 country codes. Examples: - 'California, United States' - 'Mexico' - 'US' for United States - 'DE' for Germany - 'FR' for France
output_formatNo The format of the output. Works only when parse parameter is false. - links - Most efficient when the goal is navigation or finding specific URLs. Use this first when you need to locate a specific page within a website. - md - Best for extracting and reading visible content once you've found the right page. Use this to get structured content that's easy to read and process. - html - Should be used sparingly only when you need the raw HTML structure, JavaScript code, or styling information.
user_agent_typeNoDevice type and browser that will be used to determine User-Agent header value.

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

B3.1/5.0
Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

The annotation readOnlyHint=true already indicates a safe read operation. The description adds that browser rendering and parsing are supported, which is useful. However, it does not disclose potential limitations, error behaviors, or the meaning of 'certain webpages', so it adds only modest context beyond the annotation.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is concise, with two short sentences that front-load the core purpose. It is efficient but contains a grammatical awkwardness ('a content') and vague phrasing like 'certain webpages', preventing a 5.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness2/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the presence of multiple sibling scrapers and a 5-parameter schema, this description is too sparse. It does not explain when to use this generic scraper over specialized ones like amazon_product_scraper, nor does it clarify the render or geo_location options' implications. An output schema exists, which covers return format, but selection guidance is missing.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, with detailed parameter descriptions for url, render, geo_location, output_format, and user_agent_type. The tool description adds no additional parameter meaning beyond mentioning 'different output formats', which the schema already details. Baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool gets content from any webpage and supports browser rendering, parsing, and output formats. However, it does not differentiate itself from sibling tools like ai_scraper or ai_crawler, so it falls short of a 5.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

No explicit guidance on when to use this tool versus alternatives. The mention of browser rendering and parsing hints at use cases, but there are no exclusions or comparisons to sibling scrapers, leaving the agent without clear selection criteria.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

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

  1. 6 tool updatesv0.9.2
    • Addedai_browser_agent
    • Addedai_crawler
    • Addedai_map
    • Addedai_scraper
    • Addedai_search
    • Addedgenerate_schema
  2. 6 tool updatesv0.8.1
    • Addedamazon_product_scraper
    • Addedamazon_search_scraper
    • Addedgoogle_search_scraper
    • Removedoxylabs_scraper
    • Removedoxylabs_web_unblocker
    • Addeduniversal_scraper
  3. 2 tool updatesv1.0.0
    • First observedoxylabs_scraper
    • First observedoxylabs_web_unblocker

TDQS

B3/5.0

Scored across 10 tools

Disambiguation2/5

Several tools have heavily overlapping purposes: ai_scraper and universal_scraper both claim to scrape any webpage content, while ai_crawler and ai_browser_agent both navigate websites and extract data. ai_search and google_search_scraper also cover similar territory, making selection ambiguous without very careful reading.

Naming Consistency3/5

Names are descriptive and readable, but they follow two different conventions: an ai_ prefix group (ai_crawler, ai_scraper, ai_search, ai_map, ai_browser_agent) and a target_suffix group (google_search_scraper, amazon_search_scraper, amazon_product_scraper, universal_scraper). Only generate_schema stands apart with a clear verb_noun pattern.

Tool Count5/5

Ten tools is well-scoped for a web scraping and search server covering generic scraping, search, browser automation, URL mapping, schema generation, and Amazon-specific extraction. Each tool represents a distinct product capability, even if some overlap exists.

Completeness5/5

The tool surface covers the core needs of the domain: general search, Google-specific search, generic page scraping, crawling, browser-driven interaction, site mapping, and Amazon search/product scraping. There are no obvious dead ends for common web data acquisition workflows.

Maintenance

ActivityMaintained
ResponsivenessUnresponsive

Related MCP Connectors

Related MCP Servers

  • A
    license
    Not graded
    quality
    D
    maintenance
    Enables web scraping and document processing with JavaScript execution, anti-detection measures, batch processing, and structured data extraction. Supports multiple formats including markdown, HTML, screenshots, and handles PDFs with OCR capabilities.
    4
    MIT
  • A
    license
    B
    quality
    B
    maintenance
    Enables web scraping and data extraction from websites with geographic flexibility, privacy features, and anti-detection capabilities. Supports scraping general websites, Google Search, Amazon Search, and Reddit with customizable parameters for rendering, geolocation, and locale.
    30
    404 npm
    37
    ISC
  • A
    license
    A
    quality
    B
    maintenance
    Enables AI agents to scrape any website by providing tools for JavaScript rendering, antibot bypass, and automatic captcha solving. It supports synchronous, asynchronous, and batch scraping operations with built-in proxy rotation.
    5
    207 npm
    MIT