Epstein Emails API

Query 383,000+ court-released Epstein emails via a pay-per-request API. Structured JSON. USDC on Base via the x402 protocol.

Required Credentials

This skill requires a funded EVM wallet to make payments. You must have:

| Variable | Format | Required | Description | |---|---|---|---| | XCLAW02_PRIVATE_KEY | 0x + 64 hex chars | Yes | EVM wallet private key for signing payments |

Your wallet must hold USDC on Base (eip155:8453). Each API request costs $0.001 USDC.

Security: Never log, display, or echo your private key. Use environment variables only. Use a dedicated hot wallet funded with only what you need — do not use a wallet holding significant funds.

When to Use This Skill

Use this skill when:

• The user asks about Epstein, Jeffrey Epstein, or Ghislaine Maxwell emails
• The user wants to search court-released documents or public records from the Epstein case
• The user needs structured email data (sender, recipient, date, subject, body) from the Epstein archive
• The user wants to cross-reference names, dates, or connections in the Epstein files

How to Respond

| User Says/Asks | What to Do | |---|---| | "Search Epstein emails for X" | Use /api/search — costs 1 request ($0.001). Confirm with user before calling. | | "Who emailed whom?" | Use /api/emails with from/to filters — costs 1 request ($0.001). Confirm first. | | "Show emails from a date" | Use /api/emails with date filter — costs 1 request ($0.001). Confirm first. | | "How many emails mention X?" | Use free /api/preview first (no cost). Report count. Only use paid search if user wants full results. | | "Get all emails" | Warn the user about cost first. Full pagination could cost ~$0.384 (384 requests at 1000/page). Get explicit approval before starting. | | "What's in the Epstein files?" | Explain the dataset — no API call needed. |

Important: Always confirm with the user before making paid requests. Never paginate through the full dataset without explicit user approval and a cost estimate.

Spending Safeguards

• Always estimate cost before multi-request operations. Formula: ceil(total_results / 1000) * $0.001
• Use the free /api/preview endpoint first to check result counts before committing to paid requests.
• Never auto-paginate through all results without explicit user approval.
• Single requests are fine — one search or one filtered query costs $0.001. Just confirm with the user.
• Set a spending limit in your x402 client if supported (e.g., max_amount parameter).

API Base URL

https://epsteinemails.xyz

Endpoints

GET /api/preview (FREE)

Free preview search. Use this first to check result counts before making paid requests. Rate limited (10 req/min), truncated bodies, max 10 results. No payment required.

Query Parameters:

| Param | Type | Description | |---|---|---| | q | string | Search query (min 2 characters) |

Response:

{
  "query": "american",
  "total_matches": 15,
  "returned": 10,
  "preview": true,
  "results": [
    {
      "from": "Natalia Molotkova",
      "to": "",
      "date": "Wed 2/1/2017 8:06:26 PM",
      "subject": "Round Trip ticket Barcelona/Miami",
      "body": "Title: American Express Middle seats OK? Regards, Natal...",
      "source_file": "EFTA02205655.pdf"
    }
  ]
}

GET /api/emails (PAID — $0.001)

List and filter emails with pagination. Requires x402 payment.

Query Parameters:

| Param | Type | Description | |---|---|---| | from | string | Filter by sender (case-insensitive substring) | | to | string | Filter by recipient | | subject | string | Filter by subject line | | date | string | Filter by date (e.g. "2017", "Wed") | | source_file | string | Filter by source PDF filename | | limit | int | Max results per page (default/max: 1000) | | offset | int | Pagination offset (default: 0) |

Response:

{
  "total": 383579,
  "returned": 2,
  "offset": 0,
  "limit": 2,
  "has_more": true,
  "next_offset": 2,
  "emails": [
    {
      "from": "Natalia Molotkova",
      "to": "",
      "date": "Wed 2/1/2017 8:06:26 PM",
      "subject": "Round Trip ticket Barcelona/Miami",
      "body": "Title: American Express...",
      "cc": "",
      "bcc": "",
      "source_file": "EFTA02205655.pdf",
      "source_url": "https://www.justice.gov/epstein/files/DataSet%2011/EFTA02205655.pdf"
    }
  ]
}

GET /api/search (PAID — $0.001)

Full-text search across all email fields. Requires x402 payment.

Query Parameters:

| Param | Type | Description | |---|---|---| | q | string | Search query (required, searches from/to/subject/body/date/cc/bcc) | | limit | int | Max results per page (default/max: 1000) | | offset | int | Pagination offset (default: 0) |

Response:

{
  "query": "schedule",
  "total_matches": 42,
  "returned": 2,
  "offset": 0,
  "limit": 2,
  "has_more": true,
  "next_offset": 2,
  "results": [
    {
      "index": 5,
      "email": {
        "from": "Jeffrey Epstein",
        "to": "Ghislaine Maxwell",
        "date": "Thu 3/15/2017 10:30:00 AM",
        "subject": "Schedule",
        "body": "...",
        "cc": "",
        "bcc": "",
        "source_file": "EFTA02205700.pdf",
        "source_url": "https://www.justice.gov/epstein/files/DataSet%2011/EFTA02205700.pdf"
      }
    }
  ]
}

Quick Start (Python)

# pip install "x402[httpx,evm]" eth_account

import asyncio
import os
from eth_account import Account
from x402 import x402Client
from x402.http.clients import x402HttpxClient
from x402.mechanisms.evm import EthAccountSigner
from x402.mechanisms.evm.exact.register import register_exact_evm_client

# Load private key from environment variable — never hardcode
account = Account.from_key(os.environ["XCLAW02_PRIVATE_KEY"])
client = x402Client()
register_exact_evm_client(client, EthAccountSigner(account))

async def main():
    async with x402HttpxClient(client) as http:
        resp = await http.get(
            "https://epsteinemails.xyz/api/search?q=schedule&limit=10"
        )
        data = resp.json()
        print(f"Found {data['total_matches']} matches")
        for r in data["results"]:
            e = r["email"]
            print(f"  {e['from']} -> {e['to']}: {e['subject']}")

asyncio.run(main())

Pagination

All paid endpoints support pagination. Max 1000 results per request.

Before paginating, estimate cost and get user approval:

# Step 1: Use free preview to check total matches
preview = await http.get(
    "https://epsteinemails.xyz/api/preview?q=travel"
)
total = preview.json()["total_matches"]
est_cost = ((total + 999) // 1000) * 0.001
print(f"{total} matches — full retrieval will cost ~${est_cost:.3f} ({(total + 999) // 1000} requests)")
# Step 2: Only proceed with user approval

# Step 3: Paginate
all_results = []
offset = 0
while True:
    resp = await http.get(
        f"https://epsteinemails.xyz/api/search?q=travel&limit=1000&offset={offset}"
    )
    data = resp.json()
    all_results.extend(data["results"])
    if not data["has_more"]:
        break
    offset = data["next_offset"]

Payment Details

| Field | Value | |---|---| | Protocol | x402 (HTTP 402 Payment Required) | | Price | $0.001 USDC per request | | Network | Base (eip155:8453) | | Token | USDC (0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913) | | Gas | None (facilitator-sponsored) | | Facilitator | Coinbase CDP (https://api.cdp.coinbase.com/platform/v2/x402) | | Recipient | 0xF9702D558eAEC22a655df33b1E3Ac996fAC2f1Ea |

The payment flow is automatic when using an x402-compatible client:

1. Client sends GET request
2. Server returns 402 with payment requirements in headers
3. Client signs a USDC payment and retries with payment header
4. Server verifies via Coinbase CDP facilitator, settles onchain, returns data

Data Provenance

...