Introduction

Overview

Manage Ada from your AI assistant (Claude, ChatGPT, Gemini, etc.) with the same capabilities available in the Ada dashboard.

Ada’s MCP Server connects your AI assistant directly to your Ada instance so it can analyze performance, update and create resources (knowledge articles, custom instructions, test cases), run tests, and answer natural-language questions about your Agent’s configuration and conversations. Most of the operations available in the Ada dashboard can now be done conversationally from Claude, ChatGPT, or any MCP-compatible client.

It is built on the Model Context Protocol, an open standard from Anthropic for connecting AI assistants to external data and tools.

What you can do

  • Analyze performance. Pull AR, CSAT, containment, handoffs, and conversation-level metadata. Diagnose why a metric moved, cross-reference against playbooks and coaching, and get actionable recommendations.
  • Update and create resources (knowledge articles, custom instructions, test cases). Create, update, enable, or disable them directly from your AI assistant. Changes stage safely (new knowledge is inactive by default) so you can review before they go live.
  • Run tests. Create test cases from production conversations, trigger test runs, and review pass/fail results and evaluation rationale, all without leaving your chat client.
  • Visualize trends on demand. Ask for a Sankey diagram of containment vs. resolution vs. CSAT, or a week-over-week comparison by topic. Get it.
  • Audit your setup. “Summarize our configuration” returns every playbook, action, coaching entry, and custom instruction in one view.

See the prompt library for ready-to-use prompts organized by use case.

Get started in 5 minutes

Pick your assistant and follow the walk-through:

New to MCP or not sure which client to pick? Start with the Getting started overview.

Go deeper

  • Tools — every tool the assistant can call, with parameters and examples.
  • Prompt library — use-case-driven prompts for analysis, optimization, and testing.
  • Authentication and data privacy — OAuth vs. API key, and what data flows through MCP.
  • FAQ — common questions about value, supported clients, and data handling.
  • Troubleshooting — when something goes wrong.