ARC · DEVELOPERS · MCP TOOLS · READ THE DATA

agent_leaderboard

Agent leaderboard

Cart success on ARC-100 (100 real stores, stopped before payment) for ARC browser shoppers A–E, the UCP path and the storefront MCP path, with n and 95% intervals.

NO LOGINREAD-ONLYNOT DESTRUCTIVEIDEMPOTENTNOT REACHES THE OPEN WEB

What it does

Read ARC's open agent leaderboard: cart success over the fixed ARC-100 v1 test set for each agent path ARC measures (ARC browser shoppers A-E, the UCP path, the storefront MCP path), with n, dates, evidence labels and Wilson 95% intervals, plus unverified outside submissions. Saved results only; no store requests or model calls. CC-BY-4.0.

Limits: Free saved-data read, cached 5 minutes.

Arguments

Name	Type	Required	Description
include_stores	boolean	no	Also return the per-store result grid (100 rows).

Full input JSON Schema

INPUT SCHEMA

{
  "type": "object",
  "properties": {
    "include_stores": {
      "type": "boolean",
      "description": "Also return the per-store result grid (100 rows)."
    }
  },
  "additionalProperties": false
}

Call it

From Claude, Cursor, ChatGPT or VS Code, connect the server (config) and ask in plain words; the client picks the tool. Or call it directly with JSON-RPC:

CURL · JSON-RPC

curl -s https://www.arcreport.ai/api/ask \
  -H 'content-type: application/json' \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"agent_leaderboard","arguments":{"include_stores":false}}}'

JAVASCRIPT

const res = await fetch("https://www.arcreport.ai/api/ask", {
  method: "POST",
  headers: { "content-type": "application/json" },
  body: JSON.stringify({
    "jsonrpc": "2.0",
    "id": 1,
    "method": "tools/call",
    "params": {
      "name": "agent_leaderboard",
      "arguments": {
        "include_stores": false
      }
    }
  }),
});
const { result } = await res.json();
const data = JSON.parse(result.content[0].text); // result.isError is true on errors
console.log(data);

PYTHON

import json, requests

res = requests.post("https://www.arcreport.ai/api/ask", json={
    "jsonrpc": "2.0",
    "id": 1,
    "method": "tools/call",
    "params": {
        "name": "agent_leaderboard",
        "arguments": {
            "include_stores": False
        }
    }
}, timeout=30)
result = res.json()["result"]
print(json.loads(result["content"][0]["text"]))

REST equivalent

GET /api/leaderboard. Add stores=1 for the per-store grid or format=csv for CSV. Documented in openapi.json.

CURL · GET

curl -s 'https://www.arcreport.ai/api/leaderboard'