adpilot-mcp
Enables AI agents to perform marketing operations like campaign management, content creation, and outreach, with a tiered approval model ensuring human oversight for consequential actions.
README
adpilot-mcp
The MCP (Model Context Protocol) server from AdPilot, a production system where AI agents run marketing operations, ad campaigns, content, outreach and SEO, across a set of projects, with a human in the loop for anything that matters.
This repo is the agent-facing tool surface: the MCP server and its tool definitions. The data-access layer it calls (Postgres, and the Google Ads, Meta, GA4 and Search Console integration clients) is part of the private AdPilot app and is not included here. What is published is the part worth reading: how the tools are designed, and how the approval model keeps autonomous agents safe on real accounts.
Why it exists
Agents are good at drafting a campaign, writing a post, or preparing outreach. They are not something I want spending real ad budget or emailing real people without a check. AdPilot lets agents propose actions through these tools, and a tiered approval gate decides what happens next.
The approval model
Every action-taking tool routes through queue_approval with a tier:
- Tier 1 (auto): low-risk, runs immediately (for example, pulling a performance report).
- Tier 2 (approval): anything with consequences waits for a human sign-off in Discord before it executes (publishing content, sending outreach, changing spend).
- Tier 3 (blocked): never allowed from an agent.
An agent can do the whole job end to end, but the blast radius is bounded. Read-only tools return data directly. Write tools return a queued item, not a completed action.
Tools
Read:
list_projects- projects and their integration statusget_campaign_summary- campaigns and spend for a projectget_performance- GA4 sessions and Search Console keywordsget_spend_ledger- spend history for a projectget_pending_approvals/get_pending_content/get_pending_outreach- the approval queues
Write (queued behind approval):
queue_approval- queue an action at a given tierqueue_content- queue a social post across one or more platformsqueue_outreach- queue an outreach messagecreate_blog_post/update_blog_post/publish_blog_post- blog workflowupload_asset- register a generated image, video or audio asset
Plus a project-intelligence tool group.
How it is wired
Standard MCP: a Server over StdioServerTransport, a ListToolsRequest handler that returns the catalogue above, and a CallToolRequest handler that dispatches to the tool implementations in mcp/tools/. Built on @modelcontextprotocol/sdk.
Note
This is excerpted from a live private system, for reference. It is not a runnable package on its own, the data and integration layers are omitted deliberately. It is here to show the tool design and the approval model, not to be installed.
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
Neon Database
MCP server for interacting with Neon Management API and databases
E2B
Using MCP to run code via e2b.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.