MCP Server Setup
The Buska MCP (Model Context Protocol) server lets AI assistants like Claude access your Buska data directly. It exposes search, scoring, and signal retrieval as tools that AI agents can call.
Looking for the full overview, use cases, and architecture?
See our dedicated MCP page for setup guides per client, live demos, and FAQ.
Option A: Remote HTTP MCP (recommended)
For Claude.ai web, ChatGPT Apps, and any MCP client supporting HTTP Streamable transport. OAuth handled automatically, paste a single URL:
https://api.buska.io/mcpOAuth discovery at /.well-known/oauth-authorization-server. PKCE S256 mandatory.
Option B: Local stdio (Claude Desktop, Cursor)
Add the following to your MCP client configuration file (e.g., claude_desktop_config.json for Claude Desktop):
{
"mcpServers": {
"buska": {
"command": "npx",
"args": ["-y", "buska-mcp-server"],
"env": {
"BUSKA_API_KEY": "your-api-key"
}
}
}
}Requirements
- Node.js 18 or later
- A valid Buska API key (Growth or Scale plan)
- An MCP-compatible AI client (Claude Desktop, Claude Code, or similar)
Testing the connection
After adding the configuration, restart your AI client. You should see "buska" listed as an available MCP server. Try asking: "Search Twitter for people looking for a CRM tool."
Related guides
- MCP ServerAvailable ToolsReference for the 3 MCP tools exposed by Buska. Parameters, return shapes, and example calls for AI agents.
- API ReferenceAuthenticationHow to authenticate with the Buska API. Generate an API key from Settings > API, pass it as x-api-key header. OAuth 2.1 also supported for third-party apps.
- Getting StartedWhat is BuskaBuska is a social listening platform built for B2B lead generation. It monitors 30+ social platforms for buying intent signals, scores leads with AI, and delivers qualified prospects ready for outreach.