FileMaker Data API MCP Server
Provides AI agents with direct access to FileMaker databases through the FileMaker Data API, enabling natural language interactions for querying, managing records, and database introspection.
README
FileMaker Data API MCP Server
A Model Context Protocol (MCP) server that provides AI agents with direct access to FileMaker databases through the FileMaker Data API. Compatible with Claude Desktop, Claude Code Console, Windsurf, Cursor, Cline, and other MCP-enabled AI assistants.
What It Does
This MCP server acts as a bridge between AI agents and FileMaker databases, enabling natural language interactions with your FileMaker data and database structure:
Data Interaction
- Query databases - Retrieve, search, and analyze FileMaker records
- Manage records - Create, update, delete, and duplicate records through conversation
- Execute scripts - Run FileMaker scripts with parameters
- Upload files - Add files to container fields (including repetitions)
- Set global fields - Manage FileMaker global fields programmatically
- Multi-database support - Switch between different FileMaker databases dynamically
Database Introspection
- Discover structure - AI agents can introspect database schemas, layouts, and field definitions
- Understand relationships - Access portal and related table information
- Architect solutions - Enable AI to design and implement FileMaker applications through natural language
- Metadata access - Get complete information about layouts, fields, value lists, and scripts
This introspection capability allows AI agents to understand your database architecture and help you build, modify, and optimize FileMaker solutions through simple conversation.
Quick Start
Installation
npm install -g filemaker-data-api-mcp
Setup
# Add your FileMaker connection
filemaker-mcp config add-connection production \
--server 192.168.0.24 \
--database Sales \
--user admin \
--password your_password
# Set as default
filemaker-mcp config set-default production
Configure Your AI Agent
Claude Desktop
Edit your configuration file:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"filemaker": {
"command": "filemaker-mcp",
"args": ["start"]
}
}
}
Windsurf / Cursor / Cline
Add to your MCP settings configuration:
{
"mcpServers": {
"filemaker": {
"command": "filemaker-mcp",
"args": ["start"]
}
}
}
Restart your AI agent and you're ready!
Features
Connection Management
- ✅ Multiple connections - Store and switch between different FileMaker databases
- ✅ Dynamic switching - Change databases during conversations
- ✅ Inline credentials - Connect with ad-hoc credentials without pre-configuration
- ✅ Secure storage - Credentials stored with restricted file permissions
Data Operations
- ✅ Authentication - Login, logout, session validation
- ✅ Records - Create, read, update, delete, duplicate, find
- ✅ Metadata - Access database, layout, and script information
- ✅ Container fields - Upload files (including repetitions)
- ✅ Global fields - Set global field values
- ✅ Scripts - Execute FileMaker scripts with parameters
- ✅ Portal data - Access related records through portals
Network Transport
- ✅ stdio - Local use with Claude Desktop (default)
- ✅ HTTP/HTTPS - Network deployment for remote access
Usage Examples
Example 1: Query Your Database
You ask:
"Show me all contacts from the production database"
The AI agent automatically connects and retrieves the data.
Example 2: Database Introspection
You ask:
"What is the structure of the Contacts layout? Show me all fields and their types."
The AI agent introspects the database schema and provides detailed field information.
Example 3: Switch Between Databases
You ask:
"Compare sales records from production and staging databases"
The AI agent switches between connections and compares the data.
Example 4: Create Records
You ask:
"Add a new contact: John Smith, john@example.com, 555-1234"
The AI agent creates the record with the provided information.
Example 5: Ad-hoc Connection
You ask:
"Connect to 192.168.0.26, database TestDB, user admin, password test123. Show all records."
The AI agent connects with inline credentials and queries the database.
Example 6: Solution Architecture
You ask:
"Design a customer management system with contacts, orders, and products. Create the layouts and suggest field definitions."
The AI agent uses database introspection to understand your current structure and helps architect the solution.
Configuration
CLI Commands
# Initial setup
filemaker-mcp setup
# Connection management
filemaker-mcp config add-connection <name> [options]
filemaker-mcp config remove-connection <name>
filemaker-mcp config list-connections
filemaker-mcp config set-default <name>
filemaker-mcp config show
# Start server
filemaker-mcp start
# Configure Claude Desktop
filemaker-mcp configure-claude
Environment Variables
You can also configure via environment variables in the Claude Desktop config:
{
"mcpServers": {
"filemaker": {
"command": "filemaker-mcp",
"args": ["start"],
"env": {
"FM_SERVER": "192.168.0.24",
"FM_DATABASE": "Sales",
"FM_USER": "admin",
"FM_PASSWORD": "your_password",
"FM_VERSION": "vLatest"
}
}
}
}
Configuration Priority
- Environment variables (highest) - defined in MCP config
- CLI-configured connections - stored in
~/.filemaker-mcp/config.json - Inline credentials (lowest) - passed dynamically via tool parameters
Available Tools
The MCP server provides 28 tools for interacting with FileMaker:
Authentication (3 tools)
fm_login- Authenticate with FileMaker Serverfm_logout- End sessionfm_validate_session- Check session validity
Metadata (5 tools)
fm_get_product_info- Get FileMaker Server infofm_get_databases- List available databasesfm_get_layouts- Get layouts for a databasefm_get_scripts- Get scripts for a databasefm_get_layout_metadata- Get layout field metadata
Records (7 tools)
fm_get_records- Get records with paginationfm_get_record_by_id- Get single recordfm_create_record- Create new recordfm_edit_record- Update existing recordfm_delete_record- Delete recordfm_duplicate_record- Duplicate recordfm_find_records- Search records
Container Fields (2 tools)
fm_upload_to_container- Upload file to container fieldfm_upload_to_container_repetition- Upload to repeating container field
Global Fields (1 tool)
fm_set_global_fields- Set global field values
Scripts (1 tool)
fm_execute_script- Execute FileMaker script
Configuration Tools (5 tools)
fm_config_add_connection- Add predefined connectionfm_config_remove_connection- Remove connectionfm_config_list_connections- List all configured connectionsfm_config_get_connection- Get connection details (password masked)fm_config_set_default_connection- Set default connection
Connection Tools (4 tools)
fm_set_connection- Switch to predefined connectionfm_connect- Connect with inline credentials (one-time)fm_list_connections- List available connectionsfm_get_current_connection- Show current connection details
Prerequisites
- Node.js v18 or higher
- FileMaker Server with Data API enabled
- Valid FileMaker credentials with appropriate privileges
Documentation
- Getting Started Guide - Detailed setup instructions
- User Guide - Complete usage documentation
- Token Lifecycle Management - Token caching, refresh, and error recovery
- Deployment Guide - Production deployment instructions
- API Reference - All tools documented
- Development Guide - For contributors
Security
- Credentials stored in
~/.filemaker-mcp/config.jsonwith restricted permissions (0o600) - Passwords masked in list/show commands
- Never share your config file or commit it to version control
- Use strong passwords for FileMaker Server accounts
- For production, consider using environment variables
Troubleshooting
Connection Issues
# List your connections
filemaker-mcp config list-connections
# Verify connection details
filemaker-mcp config show
Claude Desktop Not Responding
- Verify installation:
filemaker-mcp config list-connections - Rebuild if needed:
npm install -g filemaker-data-api-mcp@latest - Restart Claude Desktop
Authentication Failed
- Verify FileMaker Server is running and accessible
- Check username and password are correct
- Ensure Data API is enabled on FileMaker Server
- Verify user has proper privileges
Development
Local Development
# Clone repository (update URL after publishing)
git clone https://github.com/yourusername/filemaker-data-api-mcp.git
cd filemaker-data-api-mcp
# Install dependencies
npm install
# Build
npm run build
# Run tests
npm test
# Start in development mode
npm run dev
Use as Library
import { ConnectionManager } from 'filemaker-data-api-mcp';
const manager = new ConnectionManager();
manager.addConnection('mydb', {
server: '192.168.0.24',
database: 'Sales',
user: 'admin',
password: 'xxx',
version: 'vLatest'
});
Contributing
Contributions are welcome! See the Development Guide for implementation details and the Testing Guide for testing guidelines.
License
MIT License - see LICENSE file for details
Resources
Support
- Documentation: docs/
- Changelog: CHANGELOG.md
- Issues: GitHub Issues (update URL after publishing)
Made with ❤️ for the FileMaker and AI community
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。