nudg3-mcp
MCP server for Nudg3 brand visibility intelligence. Query your brand's AI search visibility from Claude Desktop, Cursor, Windsurf, or any MCP client.
Quickstart
1. Get an API key
Create one at app.nudg3.ai/api-keys.
Two key types (since 1.1.0):
- Workspace-scoped key — bound to one workspace. The default for single-brand customers. Every tool just works; you don't pass
workspace_id.
- Agency / company-scoped key — bound to an agency or a company that owns multiple workspaces. Use
list_workspaces first, then pass workspace_id to other tools. If the key only authorises one workspace the auto-resolver picks it for you; if it authorises many, the error response tells the LLM which UUIDs are available so it can ask you.
2. Configure Claude Desktop
Add to your Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json on Mac, %APPDATA%\Claude\claude_desktop_config.json on Windows):
{
"mcpServers": {
"nudg3": {
"command": "uvx",
"args": ["nudg3-mcp"],
"env": {
"NUDG3_API_KEY": "nudg3_live_ak_..."
}
}
}
}
3. Ask questions
- "What's my brand visibility score this week?"
- "Compare my visibility across ChatGPT and Gemini"
- "Which sources cite my brand most?"
- "Export my dashboard data for the last 30 days"
Tools
list_workspaces | read:analytics | New in 1.1.0. Portfolio discovery for agency/company keys. Workspace keys see a single-element list. |
get_overview | read:analytics | Health check: organic + branded visibility, trends, alerts, suggested investigations |
analyze_competitors | read:analytics | Organic competitive rankings, position gaps, fastest movers |
analyze_prompts | read:analytics | Prompt-level performance by funnel stage; auto-resolves your primary brand |
analyze_sources | read:analytics | Top cited domains, type distribution, content gap detection |
analyze_responses | read:analytics | Provider-grouped response samples, sentiment patterns |
query_metrics | read:analytics | Composable analytics query with measures, dimensions, filters |
get_metric_catalog | read:analytics | Discover available measures, filters, and export datasets |
export_data | export:data | CSV export for chat responses, sources, dashboard, prompts |
get_reports | read:insights | List or fetch visibility audit reports |
get_insights | read:insights | AI-generated opportunities, threats, and recommendations from a report |
get_actions | read:insights | Workspace action items with priority, effort, and impact |
Since 1.1.0, every analytics/insights/export tool accepts an optional workspace_id argument. Workspace-scoped key callers can omit it. Agency/company-scoped key callers pass the UUID returned by list_workspaces — when only one workspace is accessible, it's auto-resolved.
Environment Variables
NUDG3_API_KEY | Yes | — | Your Nudg3 API key |
NUDG3_API_URL | No | https://api.nudg3.ai | API base URL |
NUDG3_TIMEOUT | No | 30 | HTTP timeout in seconds |
Other MCP Clients
Cursor
{
"mcpServers": {
"nudg3": {
"command": "uvx",
"args": ["nudg3-mcp"],
"env": {
"NUDG3_API_KEY": "nudg3_live_ak_..."
}
}
}
}
Custom / programmatic
pip install nudg3-mcp
NUDG3_API_KEY=nudg3_live_ak_... nudg3-mcp
Development
git clone https://github.com/NUDG3-AI/nudg3-mcp-server.git
cd nudg3-mcp-server
python -m venv venv && source venv/Scripts/activate
pip install -e ".[dev]"
pytest