Convert Experiences

Convert Experiences는 Convert 계정을 AI 어시스턴트에 연결합니다. Bayesian 또는 Frequentist 설정에 따라 A/B 테스트 결과를 설명하고, 출시 전에 초안을 확인하고, 표시되지 않거나 추적되지 않는 경험의 문제를 해결하고, 테스트 프로그램을 검토하고, 저장하기 전에 승인할 수 있는 경험 초안을 만들고, Convert 문서를 검색하도록 요청할 수 있습니다. 로그인할 때 선택한 계정과 프로젝트에서만, 사용자의 Convert 권한 범위 내에서 작동하며, 항목을 삭제하려면 항상 두 번 확인해야 합니다.

convert.com에서 보기

문서

Add Convert MCP to your assistant

Use the hosted connector URL below. Your assistant will open Convert, ask you to sign in, and let you authorize selected projects or, when eligible, an entire account including future projects.

https://mcp.ai.app.convert.com/mcp

Claude, Claude Desktop, and Claude Code

  1. Open Claude settings and go to Connectors.
  2. Add a custom connector.
  3. Paste the connector URL shown above.
  4. Connect, sign in to Convert, and approve selected projects or explicitly choose an entire account.

Cursor and other remote MCP clients

  1. Create a new remote MCP server entry.
  2. Use Streamable HTTP as the transport when your client asks.
  3. Set the server URL to the connector URL shown above.
  4. Complete the OAuth flow in Convert and return to your client.

Local API-key setup

  1. Use the npm package when you need a local stdio server instead of hosted OAuth.
  2. Run npx -y @convertcom/mcp-server@latest from your MCP client configuration.
  3. Provide Convert API credentials through environment variables in that local client.

Built for real experimentation work

Ask for portfolio health, draft QA, active-test monitoring, hypothesis reviews, cleanup candidates, implementation debugging, split URL setup, and custom-code experience creation.

Secure by default

The hosted connector uses Convert OAuth. Users choose the accounts and projects they authorize, while local npm users can keep the existing API-key stdio setup unchanged.

Compact but capable

Namespaced tools keep MCP context lean while still exposing the full Convert API surface, high-level operator workflows, prompts, and OpenAI-compatible docs search.

New to Convert? Explore Convert's privacy-first experimentation platform for A/B testing, personalization, feature rollout, and optimization workflows.