AdsPilot - Google Ads MCP server for Claude: reporting + safety-gated campaign management (paused-by-default, budget caps, approval gates)
View on GitHub ↗A Google Ads MCP server that lets an AI agent manage real campaigns — without letting it spend your money unsupervised.
Connecting an LLM to an advertising account is easy. Doing it without waking up to a drained budget is the hard part. Most write-capable integrations solve this by asking the agent to confirm — which means the agent decides whether a human was consulted.
AdsPilot moves that decision out of the agent's hands. When your MCP client supports elicitation, the server asks you directly through the protocol, and the agent's own confirm flag is ignored entirely. Approval stops being a story the agent tells and becomes a fact the server can verify.
Google official MCP AdsPilot --------- Campaign writes ❌ read-only by design ✅ create, budget, keywords, ads, enable/pause Approval model n/a Human asked via MCP elicitation; agent cannot fabricate consent Fail-closed guards n/a Budget ceiling, paused-by-default, mandatory geo targeting, shared-budget protection Multi-tenant hosting ❌ self-host, single identity ✅ per-user OAuth, encrypted tokens, session isolation Site → campaign ❌ ✅ analyzesite turns any URL into campaign raw material License Apache-2.0 AGPL-3.0
This table compares Google's official server, which is deliberately read-only and therefore solves a different problem. Commercial alternatives (Markifact, Adzviser and others) do offer writes, but their internals aren't publicly auditable, so this table doesn't speculate about them. Verify current details yourself before choosing.
From the project README.
Add the radar badge to your README — it shows your project was picked up by MCP Radar and links to this page:
[](https://mcp.liqiwa.com/s/Xaena53--google-ads-mcp.html)
ChatGPT thinks. Codex works. Use ChatGPT as the planning brain while keeping the Codex harness.
0xwilliamortiz/openclaude-improvedruns anywhere. uses anything
halofyai/halofyHalofy is the open access and governance layer for AI agents across your organization. Identity, policy, provenance, audit, and signed erasure.
OpenLabs-so/openanalyticsOpen-source, privacy-first and cookieless web analytics with revenue attribution and an MCP server.
Dominic789654/awesome-deepseek-harnessA curated list of plugins, skills, MCP servers, patch/profile layers, orchestrators & UIs for DeepSeek Harness (DSH). Visualization · PPT · Coding · Agents · Loops (auto-research) and more. #dsh
pueschel88/Tradingview-MCPTradingview MCP server for controlling Tradingview Desktop from Claude Code via Chrome DevTools Protocol
The top new MCP servers of the week, every Monday. No spam, unsubscribe anytime.