How PrePilot turns your prompt into an agency workflow

After connection, you write the request naturally. PrePilot routes it through the right agency SOP, then returns a structured deliverable your team can review, edit, and use.

Ask naturally

Describe the task in your own words, just like you would brief a team member.

Match the SOP

PrePilot routes the request to the right workflow for planning, copy, SEO, reporting, audits, or campaign work.

Build the deliverable

The workflow shapes the output into clear sections, steps, checks, and ready-to-use formats.

Review and use

Your team reviews the output, edits what’s needed, then puts it into action with approval where required.

Ready to connect your AI workspace?

Choose Setup Guide

Claude Desktop Setup

  1. Open Claude Desktop -> Settings -> Connectors
  2. Click "+ Add connector" and search for custom connector option
  3. Name it "PrePilot" and enter our secure Endpoint URL
https://prepilot-system-agency.space/mcp

Claude Code Setup

  1. Add PrePilot package command to your Claude Code config
  2. Or run it directly using npx in your terminal session
CLI command
npx -y @modelcontextprotocol/server-prepilot
claude.json configuration
{
  "mcpServers": {
    "prepilot": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-prepilot"
      ]
    }
  }
}

ChatGPT Custom Connector (MCP) Integration

Add PrePilot as a custom MCP connector inside ChatGPT in under a minute. After creating the connector, you can call PrePilot from any chat.

Marketer Note: Looking for optimized strategy prompts and Sol/Terra/Luna model guides? Check out our dedicated ChatGPT Work × PrePilot Marketing Page.

  1. Open ChatGPT Connectors → Create connector. ChatGPT opens the "New App" form with the required fields.
  2. Fill the form with the values below: Name = PrePilot Agency Suite, Description = PrePilot agency workflows inside ChatGPT, Connection = Server URL (not Tunnel), Authentication = OAuth.
  3. Click Create. ChatGPT will ask you to authorize PrePilot via OAuth — sign in with your PrePilot account on app.prepilot-system-agency.space to finish.
  4. Open any chat, type @PrePilot Agency Suite followed by your request (e.g. "@PrePilot Agency Suite run an SEO audit on prepilot.ai") to invoke the connector.
PrePilot Agency Suite
PrePilot agency workflows inside ChatGPT
https://prepilot-system-agency.space/mcp

Cursor Editor Setup

  1. Open Cursor Settings -> Features -> MCP
  2. Click "+ Add New MCP Server"
  3. Choose Type: SSE and set the Server URL
prepilot
https://prepilot-system-agency.space/mcp

GitHub Copilot Integration

  1. Open VS Code / Cursor Settings -> Copilot Chat configuration.
  2. Add PrePilot MCP server with type SSE and our secure endpoint.
  3. Or use GitHub Copilot CLI to run MCP server commands.
prepilot
https://prepilot-system-agency.space/mcp

VS Code Extension Setup

Integrate PrePilot with VS Code extensions (e.g. Cline or Roo Code). Add this block to your extension's custom MCP settings file:

settings.json configuration
{
  "mcpServers": {
    "prepilot": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-prepilot"
      ]
    }
  }
}

Codex Agent Setup

  1. Open Codex Desktop Application or Agent configuration.
  2. Navigate to Plugins / Integrations -> MCP Settings.
  3. Add "PrePilot" as a custom server with our secure URL.
prepilot
https://prepilot-system-agency.space/mcp

DeepSeek Chat & API Setup

  1. Open DeepSeek Chat platform or custom API settings.
  2. Navigate to Developer -> Connected Servers.
  3. Integrate PrePilot tools using our secure SSE server address.
prepilot
https://prepilot-system-agency.space/mcp

Perplexity Pro Integration

  1. Open Perplexity settings page.
  2. Go to Connected Accounts / Custom APIs.
  3. Link PrePilot tools via our hosted OpenAPI / MCP schema link.
https://prepilot-system-agency.space/api/openapi.json

Google Gemini & Antigravity SDK Setup

Use PrePilot directly in your Google Gemini codebases or custom agents using the Antigravity (AGY) SDK:

Google AI Studio

Gemini Web & AI Studio

  1. Open Google AI Studio or Gemini Advanced custom extensions.
  2. Choose Add Custom Tool and set Name to "PrePilot".
  3. Paste our secure OpenAPI Schema URL to load all 526 skills.
https://prepilot-system-agency.space/api/openapi.json
Gemini CLI

Gemini CLI tool

  1. Install the official Gemini CLI client wrapper on your system.
  2. Append PrePilot CLI integration command to your environment profile.
  3. Run your queries directly with the --mcp option specified below.
CLI COMMAND
gemini --mcp https://prepilot-system-agency.space/mcp
Antigravity

Gemini Antigravity SDK

  1. Install the Google Antigravity (AGY) SDK from PyPI.
  2. Initialize the Agent with our secure, hosted MCP endpoint.
  3. Execute custom agent workflows with your PrePilot central balance.
Python Integration
from google_antigravity_sdk import Agent

# Initialize agent with PrePilot MCP tools
agent = Agent(mcp_servers=["https://prepilot-system-agency.space/mcp"])
result = agent.run("Create a media plan for my client")

What can you ask PrePilot MCP?

Copy these ready-to-run prompts directly into Claude, ChatGPT, or Cursor to execute marketing workflows.

SEO & Content

Write SEO Optimized Blog

Use PrePilot to write a 1500-word blog post optimized for specific keywords, including proper metadata and hooks.

Strategy & Audits

Competitor SWOT Analysis

Analyze a competitor's website URL and generate a SWOT report comparing our positioning.

Email & Funnels

Email Welcome Sequence

Generate a high-converting 3-part welcome email sequence for new signups based on target demographics.

Copywriting

Direct Response Landing Page

Draft complete direct response landing page copy for selling products with clear visual boundaries and rules.

Explore 526 Marketing Skills

Search the live catalog of skills and tools exposed to your AI assistant via our MCP endpoint.

Loading catalog...

Why integrate PrePilot via MCP?

Secure, fast, and unified marketing intelligence directly in your local workspace.

Enterprise Security

All connection streams run over secure HTTPS/SSE channels. Zero local execution exposure.

Unified Credits Billing

Execute workflows in any client (Claude, Cursor, ChatGPT) billing directly to your central PrePilot credits balance.

Zero Configuration

No complicated terminal setups. Connect via simple SSE URL or single-click Claude browser setup.

526 Marketing Skills

Access our entire catalog of battle-tested copywriting, SEO, planning, and strategy engines.

Google Cloud Run Reliability

Executed on serverless cloud environments guaranteeing high availability and fast response speeds.

Fully Bi-directional

Both the documentation and the underlying MCP tool responses natively support English and Arabic layouts.

Frequently Asked Questions

Have questions about using Model Context Protocol with PrePilot? We have answers.

Model Context Protocol is an open standard developed by Anthropic that allows AI assistants (clients) to securely connect to external databases, APIs, and local services (servers) using a standardized protocol.

Every time your AI tool calls a PrePilot workflow (e.g., runs an SEO audit or drafts copy), it consumes credits from your PrePilot account exactly as if you ran it inside our web app. You can track your usage in the PrePilot Dashboard.

Yes. PrePilot only receives the inputs required to run the specific marketing tools you invoke. We do not store or scan your private files or local directories unless you explicitly pass them as parameters to a tool.

Absolutely. PrePilot supports SSE (Server-Sent Events) which is fully compatible with Cursor's MCP settings, and can be integrated into VS Code using extensions like Cline or Roo Code.

PrePilot MCP works seamlessly with Claude (Desktop, Code, CLI), ChatGPT, Cursor, VS Code (via extensions like Cline or Roo Code), Gemini, and any other MCP-compatible client. Since it is built on the secure Server-Sent Events (SSE) standard, it integrates easily with any model or editor.

You can call all 526 marketing skills from our catalog, including SEO Blog Optimization, Competitor SWOT Analysis, Meta/Google Ads drafting, Email Sequences, and Full Copywriting blueprints. The system automatically routes your requests to the optimal AI model based on the task.

Yes. You can instruct your AI assistant to run batch content optimizations, generate multiple ad sets simultaneously, or analyze multiple competitor URLs in a single request. Folder-level content audits and bulk generation are fully supported.

Yes. You can instruct the AI assistant to save or sync the generated copies, strategies, or reports directly back to your PrePilot account or workspace, maintaining a central repository of all your assets.

The API is for developers who want to integrate PrePilot programmatically into their own apps and custom codebases with precise code-level control. The MCP is for marketers, writers, and end-users who want their AI assistants (like Claude or ChatGPT) to use PrePilot tools on their own using plain natural language, with zero coding required.

You can check your remaining PrePilot credit balance at any time (without consuming credits), search and view your generation history from the workspace, and combine PrePilot outputs with other active connectors in the same session (like drafting a copy with PrePilot and sharing it via a Slack or Gmail connector).