AFE Leaks logo

AFE Leaks

Live

Upstream well cost intelligence — AFE vs Actual costs, production, completions, surveys, formations, vendor costs, and SEC XBRL financials from 102,000+ wells.

What is AFE Leaks?

Access well-level cost data (AFE and Actual), line-item cost breakdowns, monthly production, casing designs, directional surveys, formation tops, vendor costs, and SEC financial data. Covers major U.S. basins with comprehensive filtering by operator, state, county, reservoir, and date ranges.

Available Tools

  • getCostData
    Well-level cost summaries (AFE vs Actual capex)
  • getCostBreakdown
    Line-item cost details (casing, cement, stimulation)
  • getWells
    Well metadata with dynamic column selection
  • getProduction
    Monthly production data (oil, gas, water)
  • getWellDetail
    Full well detail: metadata + costs + production
  • compareCosts
    Cost comparison across operators/basins/periods
  • searchOperators
    Search wells by operator name
  • getSurveys
    Directional survey records
  • getFormations
    Formation tops and geology
  • getVendorCosts
    Vendor cost data with breakdown
  • getFinancialNumbers
    SEC numeric financial data (10-K, 10-Q)
  • getFinancialSubmissions
    SEC company filing metadata

Requirements

  • AFE Leaks API key
    API key from AFE Leaks (register at https://afeleaks.com)
  • API or Corporate subscription
    Required for API access
  • Compatible AI agent
    Works with Claude, ChatGPT, and other MCP-compatible agents

Quick Setup Guide

Follow these steps to connect your AI agents to this connector

Sign up for PatchOps to access our MCP server management platform.

Register at https://afeleaks.com and generate an API key from your account settings.

Add your AFE Leaks API key in the PatchOps connector settings.

Connect to Your AI Assistant

Choose your AI assistant below for specific setup instructions

Follow these steps to connect AFE Leaks to Claude:

Step 1: Open Claude Settings

Launch Claude → Click the profile button (bottom left) → Navigate to Settings

Step 2: Access Connectors

Scroll down to the Connectors tab → Navigate to the bottom → Click “Add Custom Connector”

Step 3: Name Your Connector

Enter any name you'd like for this connector

Step 4: Get the Connector URL

Log in to PatchOps → Navigate to the /mcp page → Copy the connector URL you want to use

Step 5: Add the Connector

Paste the URL in the “Remote MCP Server URL” text box → Click Add

Step 6: Start Using

Your connector is now ready to use in Claude

Follow these steps to connect AFE Leaks to ChatGPT:

Step 1: Enable Developer Mode

Go to Settings → Connectors → Advanced → Enable Developer mode

Step 2: Create New Connector

Go back to Connectors page → Click “Create” (top right)

Step 3: Name Your Connector

Enter any name you'd like for this connector (descriptions are optional)

Step 4: Get the Connector URL

Log in to PatchOps → Navigate to the /mcp page → Copy the connector URL you want to use

Step 5: Add the URL

Paste the URL in the “MCP Server URL” text box

Step 6: Set Authentication

Click the authentication dropdown → Select “No authentication” (PatchOps handles authentication internally)

Step 7: Create the Connector

Click “I understand and want to continue” checkbox → Press Create

Step 8: Using the Connector

Start a new chat → Click the + button (left of text box) → Open dropdown → Hover over three dots labeled “More” → Select your connector

Instructions for connecting AFE Leaks to GitHub Copilot:

Coming Soon

Detailed setup instructions for GitHub Copilot will be added here.

For other MCP-compatible platforms:

Step 1: Get the Connector URL

Log in to PatchOps → Navigate to the /mcp page → Copy the connector URL you want to use

Step 2: Add to Configuration

Add this configuration to your MCP settings file. Replace the URL with your specific connector URL from the /mcp page:

"AFE Leaks-MCP": {
  "url": "https://patchops.ai/api/mcp/...",
  "type": "http"
}

Example Configuration

Here’s an example using PatchOps MCP:

"PatchOps-MCP": {
  "url": "https://patchops.ai/api/mcp/...",
  "type": "http"
}