Fibery

Fibery MCP Server lets you interact with Fibery data through any MCP-enabled language model provider by translating natural language into Fibery queries, reads, and mutations. It enables querying entities, describing databases, and creating or updating records conversationally.

How to use

You connect an MCP client to the Fibery MCP Server and then ask natural-language questions to read Fibery data or perform updates. Start by configuring the MCP client with the Fibery host and an API token, then use the available actions to list databases, describe a database, query data, or create and modify entities. The server handles translating your natural-language prompts into Fibery API calls and returns structured results that you can use in your workflow.

How to install

Prerequisites depend on your chosen installation method.

Option 1: Install via Smithery (recommended for Claude Desktop users)

npx -y @smithery/cli install @Fibery-inc/fibery-mcp-server --client claude

Option 2: Install via UV (local, scriptable runtime)

Prerequisites for UV: a Fibery account with an API token, Python 3.10 or higher, and the UV tool.

Install the tool using UV and then configure the MCP client to point to your Fibery workspace.

uv tool install fibery-mcp-server

Client configuration examples

In Claude Desktop, add an MCP server configuration that uses UV to run the Fibery MCP Server. The key is to specify the host and your Fibery API token so the server can access your Fibery workspace.

{
  "mcpServers": {
    "fibery-mcp-server": {
      "command": "uv",
      "args": [
        "tool",
        "run",
        "fibery-mcp-server",
        "--fibery-host",
        "your-domain.fibery.io",
        "--fibery-api-token",
        "your-api-token"
      ]
    }
  }
}

For development and testing

If you clone the server locally for development, you can run it with UV by pointing to your local clone. The command mirrors the standard UV usage and passes the host and API token.

{
  "mcpServers": {
    "fibery-mcp-server": {
      "command": "uv",
      "args": [
        "--directory",
        "path/to/cloned/fibery-mcp-server",
        "run",
        "fibery-mcp-server",
        "--fibery-host",
         "your-domain.fibery.io",
         "--fibery-api-token",
         "your-api-token"
      ]
    }
  }
}

Using the MCP server with your client

Once the MCP server is running and configured in your client, you can issue natural-language prompts to perform actions such as listing databases, describing a database structure, querying data, and creating or updating entities. The MCP server translates your prompts into Fibery API calls and returns results you can display or further process in your app.

Available actions you can perform through the MCP server

• List Databases: Retrieve all databases in your Fibery workspace.

• Describe Database: Get the structure of a database, including fields and types.

• Query Database: Execute flexible queries against Fibery data.

• Create Entity: Add a new entity with specified field values.

• Create Entities Batch: Create multiple entities in one operation.

• Update Entity: Update existing entities with new field values.

Available tools

list_databases

Retrieves a list of all databases available in your Fibery workspace.

describe_database

Provides a detailed breakdown of a specific database's structure, showing all fields with their titles, names, and types.

query_database

Offers powerful, flexible access to your Fibery data through the Fibery API.

create_entity

Creates new entities in your Fibery workspace with specified field values.

create_entities_batch

Creates multiple new entities in your Fibery workspace with specified field values.

update_entity

Updates existing entities in your Fibery workspace with new field values.