Querya

Connect AI agents to Google Ads accounts to analyze performance, perform keyword research, and safely execute campaign changes with preview and undo controls.

Querya is a hosted Model Context Protocol (MCP) server that empowers AI assistants like Claude, ChatGPT, Cursor, and Gemini CLI to interact directly with Google Ads. Through natural language conversations, agents can fetch live performance reports, run Google Keyword Planner forecasts, analyze search term waste, and generate negative keywords. Designed with safety first, Querya never directly applies mutations; instead, it validates changes with Google Ads, displays spend impacts and risk levels, and requires explicit user confirmation before applying updates, supported by per-account budget guardrails and full undo capabilities.

주요 기능

01Live performance reporting and analytics across accounts, campaigns, keywords, assets, and custom GAQL queries
02Google Keyword Planner integration for search volume, trend data, forecasts, and geo/language targeting
03Safe-by-design write operations featuring pre-change validation previews, risk assessments, and confirmation workflows
04Comprehensive change management with audit logging, guardrail caps for budgets and bids, and one-click undo support
05Hosted remote MCP architecture connecting seamlessly via Streamable HTTP and OAuth 2.1 authentication
060 GitHub stars

사용 사례

01Auditing ad performance to identify wasted spend and automatically generate negative keyword lists
02Forecasting performance, clicks, and budgets for new search or Performance Max campaigns using Keyword Planner data
03Managing daily operations such as budget adjustments, bid updates, and campaign status toggles through natural language prompts