search_enginesLABSCO SUMMARY
Discover the engine at call time instead of hard-coding 184 of them.
The schema tools mean an agent never needs the engine roster in its prompt — it searches, reads the exact parameter contract, then makes one call. Because discovery is keyless and unmetered and failed calls return zero credits, the exploration that usually costs money costs nothing here.
What it is
A remote server with five tools. Four are keyless discovery tools that find an engine and its exact input contract; the fifth runs the call and returns the result in one clean JSON envelope.
What you get
- search_engines, which finds the right engine from a plain-language need, and get_catalog, which lists every engine and category on the account
- get_engine_schema and get_action_schema, giving the actions, required and optional parameters, types and enums before a call is built
- call_engine, which runs the request and returns live data in a consistent envelope
- Rotation, residential IPs and anti-bot handling on the server side — no proxy pool to run and no per-site scraper to write
- Discovery unmetered, and calls that fail, time out or are blocked charged zero credits
Requirements
A ReefAPI key sent as a bearer token to the remote streamable HTTP endpoint. The four discovery tools work without it; only call_engine authenticates and spends, and every engine draws from the same account balance.
Setup effort
One command plus a key — claude mcp add --transport http reefapi https://api.reefapi.com/mcp --header "Authorization: Bearer YOUR_REEFAPI_KEY"
USE CASES
01
Pulling live prices for a product you are tracking
Find the right engine, read its action schema, then request the offers and buy-box for the listing.
02
Reading what people are saying about a product
Search comments on a social engine and get them back as structured records rather than scraped HTML.
03
Researching a housing or rental market
Filter a property engine by location, price and bedrooms, and get the listings back in one envelope.
04
Adding a new data source without writing a scraper
Search the catalog for what the task needs and call it the same day — no per-site parser to maintain.
5 tools
Listed in project docs. Your client may show a different list.
get_catalogget_engine_schemaget_action_schemacall_engineBEFORE YOU INSTALL
What to know first
Tools you get
5 tools named in the project's documentation.
Requires
- A client that can reach a remote MCP server over HTTP
- REEFAPI_KEY in your environment
INSTALLATION
Connect your client
Server setup
claude mcp add
Claude Code
TERMINAL
claude mcp add --transport http reefapi https://api.reefapi.com/mcp \
--header "Authorization: Bearer <YOUR_REEFAPI_KEY>"Hosted endpoint
ENDPOINT URL
https://api.reefapi.com/mcpAdd this hosted endpoint to your MCP client using the credentials listed under Requirements.
MCP Clients
Cursor
CURSOR · JSON
{
"mcpServers": {
"reefapi": {
"url": "https://api.reefapi.com/mcp",
"headers": {
"Authorization": "Bearer <YOUR_REEFAPI_KEY>"
}
}
}
}Windsurf
WINDSURF · JSON
{
"mcpServers": {
"reefapi": {
"serverUrl": "https://api.reefapi.com/mcp",
"headers": {
"Authorization": "Bearer <YOUR_REEFAPI_KEY>"
}
}
}
}ChatGPT
Settings → Connectors → add a connector by URL. Paste the endpoint and authorize with your ReefAPI key.
