MCP Server Setup
Connect AI agents to Decision Built via the Model Context Protocol. 17 tools, 8 resources, 3 prompts.
Prerequisites
- •Node.js 20 or later
- •Claude Desktop (or any MCP-compatible client)
- •A Decision Built account with an API key
1. Generate an API Key
Navigate to Settings → Integrations in your Decision Built dashboard. Create a new API key. The key format is db_live_{32 hex characters}.
Security note: The API key grants access to your organization's data. Keep it secure and never commit it to version control.
2. Configure Claude Desktop
Add the Decision Built MCP server to your Claude Desktop configuration file:
claude_desktop_config.json
{
"mcpServers": {
"decision-built": {
"command": "npx",
"args": ["@decision-built/mcp-server"],
"env": {
"DB_API_KEY": "db_live_your_key_here"
}
}
}
}The config file is located at:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
3. Restart Claude Desktop
Fully quit and relaunch Claude Desktop. The Decision Built tools will appear in the tools menu (hammer icon).
4. Test the Connection
Try these prompts in Claude Desktop:
"Show me my pipeline stages"
"List my recent contacts"
"Give me a job briefing for [contact name]"
Authorization Scopes
API keys can be configured with specific scopes to limit access:
| Scope | Access |
|---|---|
| contacts:read | Query contacts and their details |
| contacts:write | Create and update contacts |
| pipeline:read | View pipeline stages and counts |
| pipeline:write | Move contacts between stages |
| inspections:read | View inspections for contacts |
| observations:read | Access normalized findings from inspections |
| scope:read | View scope lines and suggest additions |
| pricing:read | Calculate pricing from templates |
| intelligence:read | Query learning data and patterns |
| suppliers:read | Search supplier catalogs |
| events:read | Query event history |
| projections:read | Access AI-generated projections |