Finchly - Voice-First Knowledge Sharing

Finchly MCP Server

Access your team's voice-first knowledge base directly from your AI tools via Model Context Protocol. Query insights, retrieve documentation, and connect your workflow seamlessly.

MCP Server Endpoint

https://finchly.fastmcp.app/mcp

Overview

The Finchly MCP server provides a standardized interface that allows any compatible AI model or IDE to access your team's knowledge base. Connect to our centrally hosted MCP server to query voice notes, brainstorms, and insights captured through Finchly.

Our server supports both Server-Sent Events (SSE) and standard HTTP transports, following the authenticated remote MCP specification with OAuth 2.1 authentication.

Installation Instructions

Claude Code

Native MCP support with direct authentication

Run in your terminal:

claude mcp add --transport sse finchly https://finchly.fastmcp.app/mcp

Then run /mcp in your Claude Code session to authenticate.

Claude Desktop

For Free and Pro accounts using the desktop app

1. Open configuration file:

~/Library/Application Support/Claude/claude_desktop_config.json

2. Add this configuration:

{
  "mcpServers": {
    "finchly": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://finchly.fastmcp.app/mcp"]
    }
  }
}

3. Restart Claude Desktop to apply changes

Claude.ai

For Team and Enterprise accounts on web

1. Navigate to Settings in the sidebar
2. Scroll to Integrations and click Add more
3. Enter the following:

Integration name: Finchly

Integration URL: https://finchly.fastmcp.app/mcp

4. Enable the tools in any new chats

Cursor

AI-first code editor with built-in MCP support

Option 1: Direct installation link

Install in Cursor

Option 2: Manual configuration

Visit Cursor's MCP tools page and add Finchly with URL: https://finchly.fastmcp.app/mcp

Visual Studio Code

With MCP extension installed

1. Press Cmd/Ctrl + P
2. Search for MCP: Add Server
3. Select Command (stdio)
4. Enter this command:

npx mcp-remote https://finchly.fastmcp.app/mcp

5. Name it Finchly
6. Activate via MCP: List Servers → Select Finchly → Start Server

Windsurf

AI-powered IDE from Codeium

1. Press Cmd/Ctrl + , to open settings
2. Navigate to Cascade → MCP servers
3. Select Add Server → Add custom server
4. Add this configuration:

{
  "mcpServers": {
    "finchly": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://finchly.fastmcp.app/mcp"]
    }
  }
}

Zed

High-performance collaborative editor

1. Press Cmd + , to open settings
2. Add the following configuration:

{
  "context_servers": {
    "finchly": {
      "source": "custom",
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://finchly.fastmcp.app/mcp"],
      "env": {}
    }
  }
}

Other MCP-Compatible Tools

Generic configuration for any MCP-compatible client

Command:

npx

Arguments:

-y mcp-remote https://finchly.fastmcp.app/mcp

Environment:

None

Available Tools

Once connected, your AI agent can access these Finchly tools:

search_knowledge

Query your team's knowledge base for voice notes, brainstorms, and insights

get_note

Retrieve specific voice notes or brainstorm sessions by ID

list_recent

Get the most recent knowledge entries from your team

get_tags

Browse available tags and categories in your knowledge base

Authentication

Finchly MCP server uses OAuth 2.1 with dynamic client registration for secure authentication. When you first connect from your IDE, you'll be prompted to authenticate with your Finchly account. The authentication flow is handled automatically by the MCP client.

API Key Support: For automated workflows, you can also authenticate using API keys in the Authorization: Bearer header. Contact your team admin to generate API keys for programmatic access.

Troubleshooting

Authentication errors or "internal server error"

Clear saved authentication data and try reconnecting:

rm -rf ~/.mcp-auth

Also ensure you're running Node.js 18 or later.

WSL on Windows issues

For WSL users, use this configuration instead:

{
  "mcpServers": {
    "finchly": {
      "command": "wsl",
      "args": ["npx", "-y", "mcp-remote", "https://finchly.fastmcp.app/mcp", "--transport sse-only"]
    }
  }
}

Connection timeouts or slow responses

If you experience connection issues, try using the SSE transport endpoint instead of HTTP. Replace /mcp with /sse in your configuration.