
Facebook Ads Library Mcp Server
- 848 installs
- 6 repo stars
- Updated July 18, 2026
- aradotso/marketing-skills
facebook-ads-library-mcp-server is an agent skill that documents installing and configuring a Meta Ads Library MCP server so developers can retrieve competitor ads and run batch image or Gemini video creative analysis fr
About
facebook-ads-library-mcp-server is an ara.so agent skill from the marketing-skills collection that documents how to install and configure the proxy-intell Facebook Ads Library MCP server for Claude Desktop, Cursor, and other MCP clients. The skill covers Python virtualenv setup via install.sh, ScrapeCreators and optional Gemini API keys in .env, and eight MCP tools for brand-to-platform-ID resolution, batch ad retrieval, image composition analysis, and Gemini-powered video breakdowns. Developers reach for facebook-ads-library-mcp-server when agents need live Meta competitor creative intelligence instead of guessing ad strategies—analyze_ad_videos_batch reports roughly 88% token savings versus per-video requests. The workflow supports multi-brand competitive scans, search_cached_media filters by brand and colors, credit-aware ScrapeCreators integration, and mcp.json configuration examples with troubleshooting for rate limits and cache cleanup.
- MCP server for Facebook Ads Library queries
- Batch brand lookups with caching and credit handling
- Gemini-powered video and image creative analysis
- ScrapeCreators API integration with install scripts
- Claude Desktop and Cursor MCP configuration examples
Facebook Ads Library Mcp Server by the numbers
- 848 all-time installs (skills.sh)
- +12 installs in the week ending Jul 28, 2026 (Skillselion tracking)
- Ranked #510 of 1,881 Marketing & SEO skills by installs in the Skillselion catalog
- Data as of Jul 28, 2026 (Skillselion catalog sync)
npx skills add https://github.com/aradotso/marketing-skills --skill facebook-ads-library-mcp-serverAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| Installs | 848 |
|---|---|
| repo stars | ★ 6 |
| Last updated | July 18, 2026 |
| Repository | aradotso/marketing-skills ↗ |
How do you query Meta Ads Library from an MCP agent?
Query Meta Ads Library via MCP to fetch competitor creatives and run batch image or Gemini-powered video analysis for brand themes.
Who is it for?
Developers configuring MCP in Claude Desktop or Cursor with ScrapeCreators credentials who need programmatic competitor Meta ad research.
Skip if: Teams without ScrapeCreators API access or developers who need organic SEO audits instead of paid Meta creative intelligence.
When should I use this skill?
Search Facebook ads for brands, batch-analyze competitor video creatives, or compare Meta campaigns across multiple companies.
What you get
Meta platform IDs, live ad listings with creative URLs, image and video analyses, and searchable cached media results
- MCP server configuration
- competitor ad datasets
- image and video creative analyses
By the numbers
- Documents 8 MCP tools for Ads Library queries and creative analysis
- Batch video analysis cites ~88% token savings vs single-video Gemini calls
- Default .env cache settings allow 1000 MB storage with 30-day expiry
Related skills
How it compares
Pick this skill when you need MCP-native Meta competitor creative analysis with batch Gemini video breakdowns rather than manual Ads Library browsing.
FAQ
What API keys does facebook-ads-library-mcp-server require?
facebook-ads-library-mcp-server requires a ScrapeCreators API key in .env for ad retrieval. Google Gemini is optional for analyze_ad_video and analyze_ad_videos_batch video breakdowns per the skill documentation.
How many MCP tools does the server expose?
The skill documents eight MCP tools: get_meta_platform_id, get_meta_ads, analyze_ad_image, analyze_ad_video, analyze_ad_videos_batch, get_cache_stats, search_cached_media, and cleanup_media_cache for Ads Library workflows.
Which AI clients support this MCP setup?
facebook-ads-library-mcp-server includes configuration examples for Claude Desktop claude_desktop_config.json and Cursor ~/.cursor/mcp.json, each pointing venv Python at mcp_server.py after install.sh.