API Reference

No authentication required. All endpoints return JSON. Rate limited to 60 requests/minute.

Base URL: https://agentnativeregistry.com

GET /api/tools

Search and filter all tools in the registry. Returns tools sorted by Agent Native Score descending.

Parameters

Name	Type	Description
`q`	string	Search by name, description, category, or review notes
`category`	string	Filter by category (e.g. "payments", "database", "ai")
`min_score`	number	Minimum Agent Native Score (0-100)
`has_mcp`	"true"	Only return tools with an MCP server
`limit`	number	Max results (default: 100, max: 100)

Examples

curl

JavaScript

Python

# List all tools
curl https://agentnativeregistry.com/api/tools

# Search for payment tools with MCP servers
curl "https://agentnativeregistry.com/api/tools?category=payments&has_mcp=true"

# Search by keyword
curl "https://agentnativeregistry.com/api/tools?q=email&limit=5"

// List all payment tools with MCP servers
const res = await fetch(
  "https://agentnativeregistry.com/api/tools?category=payments&has_mcp=true"
);
const { tools } = await res.json();
console.log(tools.map(t => `${t.name}: ${t.agent_native_score}/100`));

import requests

# List all payment tools with MCP servers
r = requests.get("https://agentnativeregistry.com/api/tools", params={
    "category": "payments",
    "has_mcp": "true"
})
for tool in r.json()["tools"]:
    print(f"{tool['name']}: {tool['agent_native_score']}/100")

Response

{
  "count": 3,
  "total": 126,
  "tools": [
    {
      "id": "stripe",
      "name": "Stripe",
      "agent_native_score": 58,
      "scores": { "discovery": 100, "auth": 100, "pricing": 100, "tooling": null, "reliability": null },
      "checklist": { "has_openapi_spec": true, "has_mcp_server": true, ... },
      ...
    }
  ]
}

POST /api/score

Score a tool against the checklist. Returns an existing score if the tool is in the registry, otherwise scores it on-demand using public checks only.

Request Body

Name	Type	Description
`toolName`	string (required)	Name of the tool to score
`toolUrl`	string	URL of the tool (improves accuracy)

Examples

curl

JavaScript

Python

curl -X POST https://agentnativeregistry.com/api/score \
  -H "Content-Type: application/json" \
  -d '{"toolName": "Stripe", "toolUrl": "https://stripe.com"}'

const res = await fetch("https://agentnativeregistry.com/api/score", {
  method: "POST",
  headers: { "Content-Type": "application/json" },
  body: JSON.stringify({ toolName: "Stripe", toolUrl: "https://stripe.com" })
});
const { source, tool } = await res.json();
console.log(`${tool.name}: ${tool.agent_native_score}/100 (${source})`);

import requests

r = requests.post("https://agentnativeregistry.com/api/score", json={
    "toolName": "Stripe",
    "toolUrl": "https://stripe.com"
})
data = r.json()
print(f"{data['tool']['name']}: {data['tool']['agent_native_score']}/100")

Response

{
  "source": "registry",   // or "scored" for on-demand
  "tool": {
    "id": "stripe",
    "name": "Stripe",
    "agent_native_score": 58,
    "checklist": { ... },
    ...
  }
}

MCP Server

JSON-RPC endpoint for Model Context Protocol clients. Supports search_tools, get_score, compare_tools, and list_categories.

Install

claude mcp add --transport http agent-native-registry https://agentnativeregistry.com/api/mcp

Direct JSON-RPC call

curl

JavaScript

Python

curl -X POST https://agentnativeregistry.com/api/mcp \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "id": 1,
    "method": "tools/call",
    "params": {
      "name": "get_score",
      "arguments": { "name": "Stripe" }
    }
  }'

const res = await fetch("https://agentnativeregistry.com/api/mcp", {
  method: "POST",
  headers: { "Content-Type": "application/json" },
  body: JSON.stringify({
    jsonrpc: "2.0",
    id: 1,
    method: "tools/call",
    params: { name: "get_score", arguments: { name: "Stripe" } }
  })
});
const { result } = await res.json();
console.log(result.content[0].text);

import requests

r = requests.post("https://agentnativeregistry.com/api/mcp", json={
    "jsonrpc": "2.0",
    "id": 1,
    "method": "tools/call",
    "params": {
        "name": "get_score",
        "arguments": {"name": "Stripe"}
    }
})
print(r.json()["result"]["content"][0]["text"])

Response Headers

All API responses include:

Header	Description
`X-Request-Id`	Unique UUID for this request (for debugging)
`X-API-Version`	Current API version (1.0)
`X-RateLimit-Limit`	Requests allowed per window (60)
`X-RateLimit-Remaining`	Requests remaining in current window
`X-RateLimit-Reset`	Seconds until rate limit resets

If rate limited, the response includes a Retry-After header with seconds to wait.