Claude
Claude connects to the MCP server through two clients:
- Claude Desktop — interactive analysis in a desktop app.
- Claude Code — terminal-based analysis and automation.
Both support OAuth and API key authentication and require an Ada dashboard account. Available tools depend on your role — see Permissions.
Claude Desktop
Requirements
- Claude Desktop installed.
- Node.js LTS installed locally (see recommendation).
Option 1: OAuth (recommended)
OAuth uses browser-based login with your Ada credentials. No API key is required.
-
Open Claude Desktop.
-
Go to Settings → Developer → Edit Config.
-
Add the following to your
claude_desktop_config.json: -
Replace
acme-corp.ada.supportwith your Ada instance domain (see Authentication). -
Restart Claude Desktop.
-
When prompted, authorize the connection with your Ada credentials.
Option 2: API key
API key authentication uses a token generated from your Ada dashboard. This is useful for automated setups or when browser-based login is not available.
-
Generate an API key from Config > PLATFORM > API keys in your Ada dashboard.
-
Open Claude Desktop and go to Settings → Developer → Edit Config.
-
Add the following to your
claude_desktop_config.json: -
Replace
acme-corp.ada.supportwith your Ada instance domain. -
Replace
YOUR_API_KEYwith your API key. -
Restart Claude Desktop.
Claude Code
Requirements
- Claude Code installed.
Option 1: OAuth (recommended)
Run this command, replacing acme-corp.ada.support with your Ada instance domain:
Then run /mcp inside Claude Code to authenticate in the browser.
Option 2: API key
Generate an API key from Config > PLATFORM > API keys in your Ada dashboard, then run this command, replacing the domain and key:
Run /mcp to confirm ada shows connected. Note that Claude Code stores the key in plaintext in ~/.claude.json, scoped to the directory you ran the command in — add -s user to make it available everywhere.
Next steps
- Try a prompt from the prompt library.
- Review data privacy settings.
- Connect additional AI Agents: see Connecting multiple Agents.