Skip to main content
Glama
christ0pper

extrapify

by christ0pper
OverviewSchemaRelated ServersScoreDiscussions
JavaScript
Remote

Extrapify MCP

Extract structured JSON from any public webpage inside Claude Desktop, Cursor, or any MCP-compatible client.

Define a schema. Point it at a URL. Get back validated, typed JSON.

Get an API key → extrapify.com

What this is

A thin, stateless MCP server that bridges MCP clients to the hosted Extrapify API.

Extraction does not happen inside this package. The MCP server forwards requests to the Extrapify API, which handles fetching, Browserless rendering for JS-heavy pages, Claude-powered extraction, schema validation, quota accounting, and observability on the backend.

  • MCP protocol server over stdio

  • One tool: extract_structured_data

  • Production-ready bridge, not a scraping framework

  • No extraction logic, no state, no side effects

Related MCP server: Haunt API

Install

npm install

Copy .env.example to .env and fill in your credentials:

EXTRAPIFY_API_BASE_URL=https://extrapify.com
EXTRAPIFY_API_KEY=sk_live_your_key_here

Start the server:

npm run mcp:start

Claude Desktop setup

Add this block to your Claude Desktop config (claude_desktop_config.json):

{
  "mcpServers": {
    "extrapify": {
      "command": "node",
      "args": ["/absolute/path/to/extrapify-mcp/mcp/server.mjs"],
      "env": {
        "EXTRAPIFY_API_BASE_URL": "https://extrapify.com",
        "EXTRAPIFY_API_KEY": "sk_live_your_key_here"
      }
    }
  }
}

Restart Claude Desktop. The extract_structured_data tool will appear automatically.

Cursor setup

Cursor supports stdio MCP servers. Point it at node and the local mcp/server.mjs entrypoint with the same two environment variables.

See docs/mcp-install-examples.md for copy-paste configs.

Tool: extract_structured_data

Retrieve structured JSON from any public webpage using a schema you define.

Input:

{
  "url": "https://example.com/article",
  "mode": "auto",
  "schema": {
    "title": "string",
    "author": "string",
    "published_at": "date",
    "tags": "string[]"
  }
}

Output:

{
  "extracted": {
    "title": "How Claude Agents Are Changing Developer Workflows",
    "author": "Jane Smith",
    "published_at": "2026-04-15",
    "tags": ["AI", "agents", "developer tools"]
  },
  "type": "single",
  "count": 1,
  "confidence": 0.96,
  "tokens_used": 1820
}

Supported schema types: string, number, integer, float, boolean, date, datetime, url, and any of these as arrays using [] suffix (e.g. string[]).

Supported mode values:

  • auto — let Extrapify decide based on page structure

  • single — extract the primary item only

  • list — extract all matching items as an array

Schema templates

Starter schemas for common use cases (product pages, job listings, articles, company data) are in docs/schema-templates.md.

Example workflows

Agent patterns and demo workflows are in docs/demo-workflows.md.

Other compatible clients

Any MCP client that supports stdio transport works with this package. Typically you only need:

  • command: node

  • args: absolute path to mcp/server.mjs

  • env: EXTRAPIFY_API_BASE_URL and EXTRAPIFY_API_KEY

Troubleshooting

Symptom

Fix

Server exits immediately

Verify EXTRAPIFY_API_BASE_URL is a valid absolute URL

Tool calls return 401 or 403

Check your API key at extrapify.com/dashboard

Client cannot discover tools

Confirm it is launching node against mcp/server.mjs over stdio

Requests time out

Verify the Extrapify API is reachable from your machine

JS-heavy pages return empty content

Extrapify handles Browserless fallback automatically — no action needed

Repository layout

mcp/
  server.mjs               ← MCP stdio server entrypoint
  tool-registry.mjs        ← tool definitions
  extrapify-client.mjs     ← minimal Extrapify API client
  tools/
    extract-structured-data.mjs
  configs/
    claude-desktop.local.example.json
    claude-desktop.production.example.json
docs/
  mcp-install-examples.md
  schema-templates.md
  demo-workflows.md
  mcp-marketplace-copy.md

Links

Available Tools

1 tool
extract_structured_dataA

Extract structured JSON from any public webpage using Extrapify's schema-guided extraction engine. Define the fields you want (title, price, author, tags, etc.) and their types, point the tool at a URL, and get back validated, typed JSON. Handles JavaScript-heavy pages via Browserless rendering. Ideal for scraping product pages, articles, job listings, company data, search results, and any other structured web content. Returns extracted fields, confidence score, item count, and tokens used.

ParametersJSON Schema
NameRequiredDescriptionDefault
urlYesFully qualified public webpage URL to extract structured data from (e.g. https://example.com/article). Must be publicly accessible. Does not support login-protected or paywalled pages.
schemaYesSchema definition that controls what fields to extract. Each key is the field name and each value is the field type. Supported types: "string", "number", "integer", "float", "boolean", "date", "datetime", "url", and array variants using [] suffix (e.g. "string[]"). Example: { "title": "string", "price": "number", "tags": "string[]", "published_at": "date" }. Nested objects are supported for grouped fields.
modeNoExtraction mode controlling how many items are returned. "auto" detects automatically based on page structure (recommended). "single" forces extraction of one primary item only (use for product pages, articles, profiles). "list" extracts all matching items as an array (use for search results, directories, tables). Default: "auto".auto

TDQS

A5/5.0
Behavior5/5

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

With no annotations provided, the description fully covers behavioral traits. It explains the extraction process (schema-guided, handles JS with Browserless), return format (validated typed JSON), and what is returned ('extracted fields, confidence score, item count, and tokens used'). It also discloses that it only works on public pages, setting clear expectations.

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 efficiently structured: it starts with the primary purpose, then elaborates on features and ideal use cases, and ends with return values. Every sentence adds unique information without redundancy. It is front-loaded with the most important detail (what it extracts) and fits within a few sentences.

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

Completeness5/5

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

Given the tool's complexity (3 parameters, one being a complex object schema) and no output schema, the description is remarkably complete. It covers all input parameters with usage guidance, explains the extraction engine's capabilities, and lists the output fields. No critical information is missing for an agent to decide and invoke the tool correctly.

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

Parameters5/5

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

The description adds significant value beyond the schema's property descriptions. For 'url', it specifies must be public and no login-protected. For 'schema', it provides supported types, examples, and notes nested object support. For 'mode', it explains each enum value with concrete use cases. This extra context is critical for correct usage.

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 clearly states the tool's action: 'Extract structured JSON from any public webpage' using a schema-guided engine. It specifies the verb (extract), resource (structured JSON), and context (public webpages, JavaScript-heavy handling). Since there are no sibling tools, it effectively distinguishes its purpose.

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

Usage Guidelines5/5

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

The description explicitly states when to use the tool: 'Ideal for scraping product pages, articles, job listings, company data, search results, and any other structured web content.' It also clarifies limitations: 'Does not support login-protected or paywalled pages.' The mode parameter description provides additional guidance on when to use 'single' vs 'list' modes.

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. Dates show when Glama detected each change.

  1. 1 tool updatev0.1.2
    • Changedextract_structured_data4 fields changed
      • changedInput schema / properties / mode / description
        Previous value: -"Extraction mode. Use auto unless you need a forced single object or list."New value: +"Extraction mode controlling how many items are returned. \"auto\" detects automatically based on page structure (recommended). \"single\" forces extraction of one primary item only (use for product pages, articles, profiles). \"list\" extracts all matching items as an array (use for search results, directories, tables). Default: \"auto\"."
      • changedInput schema / properties / schema / description
        Previous value: -"Schema-guided extraction definition using Extrapify field types and nested objects."New value: +"Schema definition that controls what fields to extract. Each key is the field name and each value is the field type. Supported types: \"string\", \"number\", \"integer\", \"float\", \"boolean\", \"date\", \"datetime\", \"url\", and array variants using [] suffix (e.g. \"string[]\"). Example: { \"title\": \"string\", \"price\": \"number\", \"tags\": \"string[]\", \"published_at\": \"date\" }. Nested objects are supported for grouped fields."
      • changedInput schema / properties / url / description
        Previous value: -"Public webpage URL to extract structured data from."New value: +"Fully qualified public webpage URL to extract structured data from (e.g. https://example.com/article). Must be publicly accessible. Does not support login-protected or paywalled pages."
      • changedInput schema / required
        Previous value: -[
-  "schema",
-  "url"
-]New value: +[
+  "url",
+  "schema"
+]
  2. 1 tool updatev0.1.0
    • First observedextract_structured_data
Install Server
RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

TDQS

A4.8/5.0
Disambiguation5/5

Only one tool exists, so there is no possibility of confusion or ambiguity among tools.

Naming Consistency5/5

With a single tool, naming consistency is not an issue; the name 'extract_structured_data' is clear and descriptive.

Tool Count3/5

The server has only one tool, which is thin for most domains. However, as a specialized extraction service, it is borderline acceptable.

Completeness4/5

The single tool covers the core extraction use case well, but lacks supporting features like schema management or pagination support, leaving minor gaps.

Maintenance

ActivityInactive
ResponsivenessNo issues

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Connectors

Related MCP Servers

  • webpeel by webpeel

    webpeelofficial

    webpeel
    A
    license
    Not graded
    quality
    B
    maintenance
    The web data platform for AI agents. Fetch, search, crawl, extract, monitor, and screenshot any URL. 55+ domain extractors, 65-98% token savings. 7 MCP tools included.
    332
    12
    AGPL 3.0

  • evident

    Kaushalendra-Marcus
    A
    license
    A
    quality
    B
    maintenance
    An agent-agnostic web extraction and fetch layer that turns URLs into verified, typed data with confidence scores via MCP, REST, or SDK, orchestrating scraping engines behind a resilience ladder and supporting structured extraction against any schema.
    5
    2
    Apache 2.0

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/christ0pper/extrapify-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server