About OpenWeb Ninja MCP
OpenWeb Ninja MCP is the official MCP server for OpenWeb Ninja APIs, aimed at AI agents that need real-time web data. It maps APIs across search, business, and industry data to tools, with definitions generated from OpenAPI specifications.
Key features
- 01Exposes 50 API tools
- 02Covers search, maps, jobs, and commerce
- 03Provides real estate, finance, and news data
- 04Accepts operation-specific parameters
- 05Generates tools from OpenAPI specifications
- 06Offers an on-demand subscription tool
Use cases
- 01Retrieve current web, news, or forum information
- 02Look up local businesses and map data
- 03Search jobs, products, or real estate
- 04Call OpenWeb Ninja APIs through an MCP client
MCP support
The MCP server exposes OpenWeb Ninja API operations as tools and includes a subscription tool for adding API free tiers on demand.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
npx -y @openwebninja/mcp-serverOPENWEBNINJA_API_KEYSecretYour OpenWeb Ninja API keyLocal servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.
Tools
43Saved tool list · updated 3w ago
| Parameter | Type | Required | Description |
|---|---|---|---|
| operation | enum | Yes | Which endpoint to call. See the tool description for each operation and its parameters.ai_overviews |
| args | object | No | Parameters for the chosen operation as key/value pairs (see the tool description for required params). |
Details
From the project READMEOpenWeb Ninja MCP Server
Official Model Context Protocol server for OpenWeb Ninja APIs. Gives any MCP-compatible AI agent (Claude, Cursor, Cline, and others) real-time access to web search, local business data, jobs, e-commerce, real estate, finance, news, and more, through a single connection.
One tool per API product (50 tools), each exposing that API's operations, plus a subscribe tool for adding an API's free tier on demand — 51 tools in total. Schemas are generated directly from OpenWeb Ninja's OpenAPI specs, so the server always matches the live APIs.
Setup
You need an OpenWeb Ninja API key. Get one at openwebninja.com.
The server runs via npx — there's nothing to install globally. Pick your client below; npx fetches @openwebninja/mcp-server on demand.
Claude Desktop
Add to claude_desktop_config.json (Settings -> Developer -> Edit Config):
{
"mcpServers": {
"openwebninja": {
"command": "npx",
"args": ["-y", "@openwebninja/mcp-server"],
"env": { "OPENWEBNINJA_API_KEY": "your-api-key" }
}
}
}
Claude Code
claude mcp add openwebninja -e OPENWEBNINJA_API_KEY=your-api-key -- npx -y @openwebninja/mcp-server
Cursor / Cline / Continue / Windsurf
Use the same command / args / env shape in the client's MCP config (mcp.json or equivalent).
From source (local development)
git clone <repo> && cd openwebninja-mcp
npm install
npm run build
Then point your MCP client at the built entry:
{
"mcpServers": {
"openwebninja": {
"command": "node",
"args": ["/absolute/path/to/openwebninja-mcp/dist/index.js"],
"env": { "OPENWEBNINJA_API_KEY": "your-api-key" }
}
}
}
How the tools work
Each tool maps to one OpenWeb Ninja API and takes two inputs:
operation: which endpoint to call (e.g.search,product_details)args: the parameters for that operation
The tool description lists every operation and its required parameters. Example call to the jsearch tool:
{
"operation": "search",
"args": { "query": "site reliability engineer remote", "country": "us" }
}
Arguments are validated and type-coerced before the request is sent, and responses return the OpenWeb Ninja data payload plus a request_id.
Environment variables
| Variable | Required | Description |
|---|---|---|
OPENWEBNINJA_API_KEY | yes | Your OpenWeb Ninja API key (sent as x-api-key). |
OPENWEBNINJA_BASE_URL | no | Override the API host (defaults to https://api.openwebninja.com). For staging/testing. |
Available tools (51)
Search & discovery: realtime_web_search, realtime_bing_search_data, realtime_news_data, real_time_news_search, realtime_forums_search, web_search_autocomplete, realtime_image_search, reverse_image_search, realtime_lens_data, real_time_video_search, realtime_shorts_search, ai_overviews, google_ai_mode, social_links_search
Local & maps: local_business_data, local_business_database, yelp_business_data, trustpilot_company_and_reviews, local_rank_tracker, driving_directions, waze, google_maps_traffic_alerts, ev_charge_finder
Jobs & companies: jsearch, realtime_linkedin_data, realtime_indeed_data, realtime_ziprecruiter_data, realtime_monster_data, realtime_bing_jobs_data, job_salary_data, realtime_glassdoor_data
Commerce & product: realtime_amazon_data, realtime_product_search, real_time_walmart_data, real_time_ebay_data, realtime_costco_data, real_time_wayfair_data, realtime_homedepot_data, realtime_books_data, play_store_apps
Real estate: realtime_zillow_data, real_time_redfin_data
Finance & events: realtime_finance_data, realtime_events_data
Contact & enrichment: website_contacts_scraper, email_search
Utility: web_unblocker
LLM relays: chatgpt, gemini, copilot
Access: subscribe (add an API's free tier on demand)
Development
npm run sync # sync OpenAPI specs from S3 into openapi-cache/ (needs AWS creds)
npm run generate # regenerate src/generated/manifest.ts from the specs
npm run build # compile TypeScript to dist/
npm run dev # run the server over stdio (tsx, no build)
npm run inspect # launch the MCP Inspector against the server
npx tsx test/smoke.ts # smoke test: list tools + validation + request path
The pipeline is: OpenAPI specs (openapi-cache/) -> scripts/generate.ts -> src/generated/manifest.ts -> tools registered at runtime in src/server.ts. To add or update an API, re-sync the specs and regenerate; no hand-written per-tool code is needed (only the curated descriptions in src/lib/descriptions.ts).
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.