Awesome MCP › Other MCP Servers

gomarble-ai/facebook-ads-mcp-server

⭐ 365 Python added to this list on 2025-05-06 repository created 2025-04-07

facebook-ads-mcp-server is a Model Context Protocol server written in Python that acts as an interface to the Meta Ads (Facebook Ads) platform, giving MCP-compatible assistants programmatic access to advertising data and management features. The server is launched as a stdio process and receives a Meta User Access Token through the --fb-token command-line argument, so it is configured in a client such as Cursor or Claude Desktop by pointing the client at server.py and supplying the token. The project also ships one-click installers that set up the server locally, handle the environment and, when no Meta developer token is supplied, obtain a token through the maintainer's hosted service; the README states the token is stored locally rather than on that service. The tool surface is organised into four groups. Account and object reads cover list_ad_accounts, get_details_of_ad_account and per-object lookups for campaigns, ad sets, ads and ad creatives, including a batch call for multiple ad sets. Collection fetches walk the account hierarchy: campaigns, ad sets and ads by ad account, ad sets and ads by campaign, ads by ad set, and creatives by ad. Insight tools return performance data at ad account, campaign, ad set and ad level, with fetch_pagination_url for following paged result URLs. Activity tools return change history for an ad account or an ad set. Most tools accept extra parameters such as fields, filtering, limit, pagination cursors and date ranges. Requirements are Python 3.10 or newer plus the mcp and requests packages, and the server can also be installed through Smithery. It is aimed at marketers and developers who want conversational or agent-driven reporting over Meta advertising accounts. Licensed under MIT.

https://github.com/gomarble-ai/facebook-ads-mcp-server

mcp-serverpythonfacebookmeta-adsadvertisingmarketinganalyticsapi-integration