Pipeboard
Run ads on five platforms by chatting with your AI assistant
About
Pipeboard is a connector layer that plugs your Meta (Facebook/Instagram), Google, TikTok, Snap and Reddit ad accounts into the AI assistant you already use — Claude, ChatGPT, Cursor, Claude Code, n8n or any other MCP client. Once an account is authorized, you just say "raise this ad set budget to 500", "clone last week's best campaign" or "upload these images to the creative library", and the assistant carries it out on the platform instead of you clicking through Ads Manager. Unlike ad tools that ship their own UI and strategy agent, it produces no strategy and no reporting dashboard: it hands the platforms' write capability to your assistant (the vendor lists 120+ write tools, 60+ on Meta and 35+ on Google) behind three built-in guardrails — new campaigns and ads are created paused, tokens can be read-only, and tokens can be scoped to specific ad accounts. It holds Meta Business Partner status and production-level (Standard) Google Ads API access, and its Meta MCP server, meta-ads-mcp, is open source on GitHub. The vendor also states it manages over $400M in monthly ad spend across 10,000+ connected ad accounts; those two figures are vendor-reported and not independently verified.
Key Features
- ✓Connect five ad platforms: Meta, Google, TikTok, Snap and Reddit
- ✓Create, clone and bulk-edit campaigns, ad sets and ads in natural language
- ✓Upload images and video into the platform creative library
- ✓Query spend, conversions and ROAS with drill-down by level
- ✓Create campaigns and ads paused by default, with read-only or account-scoped tokens
- ✓Partner program: one API key manages unlimited client ad accounts
Who it's for
- In-house media buyers and growth leads who tweak budgets, clone campaigns and ship creatives daily
- Agencies and freelance media consultants juggling many client ad accounts
- Not for teams without an AI assistant subscription, or those who depend on a fixed cross-platform reporting dashboard
Pros & cons
Pros
- Runs inside the Claude, ChatGPT or Cursor you already pay for, with no new ads UI to learn
- Goes further than the mostly read-only ad MCPs, with 60+ Meta and 35+ Google tools
- Guardrails are product defaults rather than user discipline — paused-by-default creation, read-only or account-scoped tokens
- Meta Business Partner status and Standard Google Ads API access back the integrations' stability
Limitations
- It is only a connector, so you must bring your own Claude or ChatGPT style subscription and pay model costs separately
- The free tier is capped at 30 calls per week and 2 account slots, and extra accounts cost $10/mo each, so agency-scale usage grows cost linearly
- There is no reporting dashboard of its own, so cross-platform comparison means re-asking the assistant — a poor fit for teams needing fixed daily reports
- It lets a model write to live ad accounts; paused-by-default creation does not cover budget edits or bulk changes, so you still need your own review step