Agent1st Ads MCP

What Nolas-Shadow/agent1st-ads-mcp MCP server does

Nolas-Shadow/agent1st-ads-mcp MCP server exposes advertising operations for Meta and TikTok through the Model Context Protocol. It is intended for AI agents that need to create campaigns and inspect their results without manually coordinating each platform’s individual campaign objects.

For Meta, the server supports campaigns, ad sets, and ads across Facebook and Instagram. For TikTok, it supports campaign creation. The repository describes campaign creation as a single-call workflow that can include targeting, creative, budget, and the ad. This makes the server relevant when an agent should assemble the main campaign inputs as one operation instead of issuing a long sequence of platform API calls.

The server also provides performance retrieval. The documented quick-start flow names get_meta_campaign_stats and get_tiktok_campaign_stats for checking campaign results after launch.

How it works

The account connection uses OAuth. A user first connects an advertising account, then the agent can invoke the applicable campaign tool. The README does not describe the OAuth consent screens, token storage, account-selection behavior, or the detailed input schema, so those implementation details should be confirmed in the provider’s full documentation.

The main launch operations are named create_meta_campaign and create_tiktok_campaign. Meta operations map to Facebook and Instagram advertising, while TikTok operations target TikTok campaigns. After creation, statistics tools provide current campaign performance data according to the repository’s feature description.

Nolas-Shadow/agent1st-ads-mcp MCP server is therefore a fit for workflows that combine campaign execution with agent-led monitoring. It is not documented here as a general social publishing tool, analytics platform for unrelated channels, or ad creative generation system.

Setup and configuration

Using Nolas-Shadow/agent1st-ads-mcp MCP server requires a license from Agent 1st Ads. The README lists paid plans: Starter supports one platform, Pro covers Meta and TikTok, Premium adds AI optimization, and Elite adds white-label use and unlimited access. The listed prices are $29 per month, $69 per month, $149 per month, and $399 per month respectively.

The setup sequence documented by the project is short: connect the ad account through OAuth, call the platform-specific campaign creation tool, and use the corresponding statistics tool to monitor performance. Full documentation is hosted at agent1st.io/ads. No environment variable names, local package command, or configuration file format is provided in the supplied material, so an installation command cannot be specified reliably here.

Tools and capabilities

The documented operations are:

• create_meta_campaign for creating Meta campaigns.
• create_tiktok_campaign for creating TikTok campaigns.
• get_meta_campaign_stats for retrieving Meta campaign performance data.
• get_tiktok_campaign_stats for retrieving TikTok campaign performance data.

The broader feature list also identifies Meta campaign, ad set, and ad management, along with TikTok campaign management. The exact parameters, response shapes, supported objective types, creative formats, budget constraints, and update operations are not included in the excerpt. Consult the provider documentation before building a production workflow around those details.

Limitations and notes

The server is a licensed commercial service rather than a repository-only, no-cost integration. Its documented platform coverage is limited to Meta and TikTok. OAuth account connection is required, but the supplied material does not specify which OAuth provider screens or permissions are involved.

The README says that real-time statistics are available, but it does not define refresh behavior, latency, retention, or the metrics returned. It also does not provide a local installation command or identify compatible desktop clients. Confirm deployment and client configuration with Agent 1st Ads before implementation. Support is listed at solutions@agent1st.io.