Skip to main content
Want your AI agent to set it up? Copy one prompt or install the Spyro AI skill: Set up with your AI agent.
Every app uses the same server address:
Choose how to sign in:
  • Sign in - the app opens a Spyro AI page where you approve access. Use this whenever the app supports it.
  • Key - you create an MCP key and the app sends it in a header. See Manage keys and apps to create one.
Each of the eight main apps has a full guide with tips and troubleshooting:

Claude

Claude Code

ChatGPT

Gemini

Gemini CLI

Codex

Cursor

VS Code

AI apps change their menus often. If a button has moved, search your app’s settings for “connectors”, “MCP” or “custom connector”, or check the app’s own documentation.

Claude (claude.ai and Claude Desktop)

1

Add a custom connector

Open Customize → Connectors, click + Add, then Add custom connector.
2

Enter the address

Name it Spyro AI and paste the server address. Leave the advanced OAuth fields empty.
3

Connect

Click Add, then Connect. Sign in to Spyro AI, pick your workspaces and click Allow.
On Team and Enterprise plans an owner adds the connector first in Organization settings → Connectors (Add → Custom → Web). Members then click Connect.

Claude Code

Then run /mcp inside Claude Code, pick spyro-ai, and sign in in your browser.

ChatGPT

1

Open the connector settings

Go to chatgpt.com/plugins, click the plus button, then Add custom MCP server.
2

Enter the address

Name it Spyro AI, paste the server address under Connection and choose OAuth.
3

Accept and sign in

Read the risk notice, click I understand and want to continue, then Create as a plugin.
4

Install and sign in

Install the plugin from your plugins list and sign in to Spyro AI.
Custom MCP servers in ChatGPT depend on your plan and workspace settings. On Business, Enterprise and Education workspaces an admin may need to allow them first.

Gemini

1

Open Developer Mode

Go to gemini.google.com/devmode (also in Settings → Gemini Apps Developer Mode).
2

Create the server

Click + Create new and paste the server address as the MCP Server URL. Leave the client settings empty.
3

Connect

Accept the confirmation, click Connect, name the app Spyro AI and sign in.
Gemini Enterprise needs its own OAuth client.
  1. In Spyro AI open Settings → MCP → OAuth clients and create a client with this redirect URI: https://vertexaisearch.cloud.google.com/oauth-redirect. Copy the client ID and secret.
  2. In the Google Cloud console open Gemini Enterprise → Apps → your app → Connected data stores → New data store, search for Custom MCP Server and click Add MCP server.
  3. Choose OAuth 2.0, enter these values plus your client ID and secret, and turn on Enable PKCE Support (Spyro AI requires it):
  1. Finish the steps and turn the connector on. For Read and take actions, use the scopes spyro.read spyro.actions. If the Custom MCP Server option is missing, a Google Cloud organization policy may block it.

Gemini CLI

Add the server to ~/.gemini/settings.json. The httpUrl field selects the streamable HTTP transport:
~/.gemini/settings.json
Then run /mcp auth spyro-ai inside Gemini CLI and sign in in your browser.

Codex

Add the server to ~/.codex/config.toml:
~/.codex/config.toml
Then sign in:

Cursor

Add the server to mcp.json (project .cursor/mcp.json or global ~/.cursor/mcp.json). Cursor asks you to sign in on first use.
mcp.json

VS Code (GitHub Copilot)

Add the server to .vscode/mcp.json. VS Code asks you to sign in.
.vscode/mcp.json

Other apps

Any other app

Apps that speak MCP over streamable HTTP: paste the address and sign in. Apps without sign-in can send a key as Authorization: Bearer sk_mcp_... or X-API-Key: sk_mcp_.... Apps that only run local (stdio) servers can use mcp-remote, which opens your browser to sign in:
Never put a key in the server address or a query string. Spyro AI only reads keys from the Authorization or X-API-Key header. Keep keys in an environment variable, not in a file you commit.

Check that it works

Ask your app: “List the Spyro AI tools you can use.” then “Use Spyro AI to give me an overview of my workspace.” If you connected more than one workspace, the app first asks which one to use. If something fails, see Troubleshooting. Building your own client or agent? See the Developer guide.