flour-ads
MCP server for managing ad campaigns across Google Ads, Meta, and more. Enables deploying campaigns, checking performance, and managing budgets from terminal or AI assistants.
README
flour-ads
CLI + MCP server for managing ad campaigns across Google Ads, Meta, and more.
Deploy campaigns, check performance, and manage budgets — from the terminal or through any MCP-compatible AI assistant (Claude Code, Cursor, etc.).
Install
git clone https://github.com/Wayy-Research/flour-ads.git
cd flour-ads
uv venv && source .venv/bin/activate
uv pip install -e ".[dev]"
Quick Start
CLI
# List available platforms
flour-ads platforms
# Authenticate (uses env vars or stored credentials)
flour-ads auth login --platform google_ads --account-id 123-456-7890
# Deploy a campaign from YAML config
flour-ads campaign create \
--platform google_ads \
--account-id 123-456-7890 \
--config examples/ringgold_campaign.yaml
# List campaigns
flour-ads campaign list --platform google_ads --account-id 123-456-7890
# Check performance
flour-ads campaign performance \
--platform google_ads \
--account-id 123-456-7890 \
--campaign-id 12345678 \
--range LAST_7_DAYS
# Pause/resume
flour-ads campaign pause --platform google_ads --account-id ID --campaign-id CID
flour-ads campaign resume --platform google_ads --account-id ID --campaign-id CID
# Update budget
flour-ads budget update \
--platform google_ads \
--campaign-id 12345678 \
--account-id 123-456-7890 \
--daily-budget 45.00
MCP Server
# Start the MCP server (stdio transport)
flour-ads serve
Add to your Claude Code settings (.claude/settings.json):
{
"mcpServers": {
"flour-ads": {
"command": "flour-ads",
"args": ["serve"]
}
}
}
MCP Tools:
| Tool | Description |
|---|---|
list_platforms |
List available ad platforms |
authenticate |
Authenticate with a platform |
create_campaign |
Create a campaign from config file or inline params |
pause_campaign |
Pause an active campaign |
resume_campaign |
Resume a paused campaign |
list_campaigns |
List campaigns for an account |
update_budget |
Update daily budget |
get_performance |
Get performance metrics with daily breakdown |
Campaign Config (YAML)
Define campaigns as YAML files for reproducible, version-controlled deployments:
name: "My Search Campaign"
platform: google_ads
campaign_type: SEARCH
status: PAUSED # deploy paused, review, then enable
budget:
daily_amount: 50.00
bidding:
strategy: MAXIMIZE_CONVERSIONS
targeting:
locations: ["10001", "10002", "10003"]
languages: ["en"]
networks:
search: true
display: false
ad_groups:
- name: "Core Keywords"
keywords:
- text: "plumber NYC"
match_type: PHRASE
ads:
- headlines:
- "NYC Plumber — Fast Service"
- "Licensed & Insured"
- "Free Estimates Today"
descriptions:
- "Trusted plumber serving NYC. Same-day service. Call now."
- "20+ years experience. No hourly rates. Upfront pricing."
final_url: "https://example.com"
negative_keywords:
- "DIY"
- "jobs"
- "hiring"
See examples/ for a complete real-world campaign config.
Credentials
flour-ads looks for credentials in this order:
- OS Keyring (macOS Keychain, Linux Secret Service, Windows Credential Locker)
- JSON file at
~/.flour-ads/credentials.json - Environment variables
Google Ads
export GOOGLE_ADS_DEVELOPER_TOKEN="..."
export GOOGLE_ADS_CLIENT_ID="..."
export GOOGLE_ADS_CLIENT_SECRET="..."
export GOOGLE_ADS_REFRESH_TOKEN="..."
export GOOGLE_ADS_CUSTOMER_ID="1234567890"
See docs/google-ads-setup.md for the full setup guide.
Adding a Platform
flour-ads uses a plugin architecture. To add a new platform:
- Create
src/flour_ads/platforms/yourplatform/adapter.py - Inherit from
AdPlatformand implement all abstract methods:
from flour_ads.platforms.base import AdPlatform
class YourPlatformAdapter(AdPlatform):
platform_id = "yourplatform"
platform_name = "Your Platform"
async def authenticate(self, account_id, credentials): ...
async def create_campaign(self, account_id, config): ...
async def pause_campaign(self, account_id, campaign_id): ...
async def resume_campaign(self, account_id, campaign_id): ...
async def get_performance(self, account_id, campaign_id, date_range): ...
async def list_campaigns(self, account_id, status=None): ...
async def update_budget(self, account_id, campaign_id, daily_budget): ...
- Register via entry point in
pyproject.toml:
[project.entry-points."flour_ads.platforms"]
yourplatform = "flour_ads.platforms.yourplatform.adapter:YourPlatformAdapter"
Platforms
| Platform | Status | Notes |
|---|---|---|
| Google Ads | Ready | Search campaigns, RSA ads, keyword targeting |
| Meta (Facebook/Instagram) | Stub | Adapter skeleton in place |
| TikTok | Planned | |
| Planned | ||
| Microsoft Ads | Planned |
Development
uv pip install -e ".[dev]"
pytest tests/ -v
black . && ruff check .
mypy src/
License
MIT — see LICENSE.
Built by Wayy Research — Buffalo, NY
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.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
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.