float-mcp
A community MCP server for float.com.
README
Float MCP Server
A comprehensive Model Context Protocol (MCP) server that provides seamless integration with Float.com - the resource management and project planning platform. This server exposes Float's complete API as MCP tools, enabling AI assistants like Claude to interact with Float for project management, resource allocation, time tracking, and team coordination.
🌟 Features
Complete Float API Coverage
- 👥 People Management - Full CRUD operations for team members
- 📁 Project Management - Projects, phases, tasks, and milestones
- ⏱️ Time Tracking - Logged time, timesheets, and billable hours
- 📊 Resource Allocation - Team member assignments and scheduling
- 🎯 Task Management - Project tasks, dependencies, and workflows
- 🏢 Organization - Departments, roles, accounts, and permissions
- 📅 Time Off - PTO requests, approvals, and holiday management
- 📈 Reports - Time reports, utilization, and project analytics
Advanced Features
- 🔄 Rate Limiting - Built-in API rate limiting with exponential backoff
- 🛡️ Type Safety - Full TypeScript support with Zod schema validation
- 📝 Comprehensive Logging - Detailed logging for debugging and monitoring
- ⚡ Performance - Optimized for fast response times and efficient API usage
- 🧪 Testing - Comprehensive integration test suite
- 🐳 Docker Support - Ready-to-deploy Docker container
🚀 Quick Start
Prerequisites
- Node.js 22.0.0 or later
- Float.com account with API access
- Valid Float API key
1. Installation Options
A. DXT Extension (Recommended)
For DXT-compatible environments, download the latest DXT package:
📦 Download Float MCP Extension (.dxt)
Install the .dxt file through your DXT-compatible application.
B. Manual Installation
# Using npm
npm install
# Using yarn
yarn install
2. Configuration
Create a .env file in the project root:
# Float API Configuration
FLOAT_API_KEY=your_float_api_key_here
FLOAT_API_BASE_URL=https://api.float.com/v3
# Optional: Enable debug logging
LOG_LEVEL=info
3. Build & Run
# Build the project
npm run build
# Start the MCP server
npm start
# For development with auto-reload
npm run dev
4. MCP Client Integration
For DXT Installation
After installing the DXT extension, configure your Float API key in the extension settings. The extension will handle the MCP server configuration automatically.
For Manual Installation
Add to your MCP client configuration (e.g., Claude Desktop):
{
"mcpServers": {
"float-mcp": {
"command": "docker",
"args": [
"run",
"--rm",
"-i",
"-e",
"FLOAT_API_KEY=YOURAPIKEY",
"-e",
"LOG_LEVEL=debug",
"ghcr.io/asachs01/float-mcp:latest"
]
}
}
}
🚀 Available Tools
The Float MCP server provides 4 optimized decision-tree tools that efficiently replace 246+ granular tools while maintaining complete functionality:
🔧 Optimized Tools (Recommended)
1. manage-entity - Core Entity Management
Consolidates all CRUD operations for core entities with decision tree routing:
manageEntity({
entity_type: "people" | "projects" | "tasks" | "clients" | "departments" | "roles" | "accounts" | "statuses",
operation: "list" | "get" | "create" | "update" | "delete" | "get-current-account" | "bulk-update-account-permissions",
// ... entity-specific parameters
})
Replaces: All CRUD tools for people, departments, roles, accounts, projects, tasks, clients, and statuses (~120 tools)
2. manage-project-workflow - Project Operations
Handles all project-specific workflow operations:
manageProjectWorkflow({
workflow_type: "phases" | "milestones" | "project-tasks" | "allocations",
operation: "list" | "get" | "create" | "update" | "delete" | "complete" | "archive" | "bulk-create" | "reorder",
// ... workflow-specific parameters
})
Replaces: Project phases, milestones, project tasks, allocations, dependencies, bulk operations (~60 tools)
3. manage-time-tracking - Time Management
Manages all time-related operations with comprehensive reporting:
manageTimeTracking({
tracking_type: "logged-time" | "timeoff" | "public-holidays" | "team-holidays",
operation: "list" | "get" | "create" | "update" | "delete" | "approve" | "reject" | "bulk-create",
report_type?: "person-summary" | "project-summary" | "timesheet" | "billable-analysis",
// ... time-specific parameters
})
Replaces: Logged time, time off, holidays, approvals, timesheets, summaries (~45 tools)
4. generate-report - Analytics & Reporting
Comprehensive reporting and analytics engine:
generateReport({
report_type: "time-report" | "project-report" | "people-utilization-report" | "capacity-report" | "budget-report",
// Advanced filtering and grouping options
group_by?: "person" | "project" | "client" | "department" | "date" | "week" | "month",
include_details?: boolean,
// ... extensive reporting parameters
})
Replaces: All reporting tools with advanced analytics, grouping, filtering (~20 tools)
📊 Optimization Benefits
- 🔥 Massive Efficiency: 246+ tools → 4 optimized tools (98.4% reduction)
- 🧠 AI-Friendly: Decision tree parameters instead of tool proliferation
- ⚡ Better Performance: Consolidated API calls and reduced overhead
- 🔒 Full Compatibility: Zero functionality loss, complete backward compatibility
- 🛠️ Easier Maintenance: Centralized logic with consistent patterns
🔄 Legacy Tools (Backward Compatibility)
All original 246+ granular tools remain available for backward compatibility:
Reporting & Analytics
- Reports:
get-time-report,get-project-report,get-people-utilization-report - Analytics:
get-person-logged-time-summary,get-project-logged-time-summary
🔧 Configuration
Environment Variables
| Variable | Description | Required | Default |
|---|---|---|---|
FLOAT_API_KEY |
Your Float API key | ✅ Yes | - |
FLOAT_API_BASE_URL |
Float API base URL | ❌ No | https://api.float.com/v3 |
LOG_LEVEL |
Logging level (error, warn, info, debug) |
❌ No | info |
MAX_RETRIES |
Maximum API retry attempts | ❌ No | 3 |
REQUEST_TIMEOUT |
API request timeout (ms) | ❌ No | 30000 |
Getting a Float API Key
- Log in to your Float account
- Go to Settings > API > Personal Access Tokens
- Click Generate New Token
- Copy the API key
- Add it to your
.envfile
📖 Usage Examples
Basic Operations
// List all projects
const projects = await listProjects({});
// Get a specific project
const project = await getProject({ project_id: 12345 });
// Create a new person
const person = await createPerson({
name: 'John Doe',
email: 'john@example.com',
department_id: 1,
});
// Schedule an allocation
const allocation = await createAllocation({
project_id: 12345,
people_id: 67890,
start_date: '2024-01-15',
end_date: '2024-01-31',
hours: 8,
});
Advanced Workflows
// Get team utilization report
const utilization = await getPeopleUtilizationReport({
start_date: '2024-01-01',
end_date: '2024-12-31',
});
// Bulk create project tasks
const tasks = await bulkCreateProjectTasks({
project_id: 12345,
tasks: [
{ name: 'Design Phase', start_date: '2024-01-01' },
{ name: 'Development Phase', start_date: '2024-01-15' },
],
});
// Process time off request
await createTimeOff({
people_ids: [67890],
timeoff_type_id: 1,
start_date: '2024-02-01',
end_date: '2024-02-05',
status: 1, // Pending approval
});
await approveTimeOff({
timeoff_id: 123,
approved_by: 456,
notes: 'Approved for vacation',
});
🧪 Testing
Running Tests
# Run all tests
npm test
# Run integration tests (requires API key)
npm run test:integration
# Run tests with coverage
npm run test:coverage
# Run specific test suites
npm run test:integration:mock # Mock API responses
npm run test:integration:real # Real API calls (use with caution)
Test Configuration
For integration tests with real API calls:
# .env.test
FLOAT_API_KEY=flt_your_test_api_key
TEST_REAL_API=true
TEST_MOCK_MODE=false
🐳 Docker Support
Using Docker
# Build the image
docker build -t float-mcp .
# Run the container
docker run -d \
--name float-mcp \
-e FLOAT_API_KEY=your_float_api_key_here \
-p 3000:3000 \
float-mcp
Docker Compose
version: '3.8'
services:
float-mcp:
build: .
environment:
- FLOAT_API_KEY=your_float_api_key_here
- LOG_LEVEL=info
ports:
- '3000:3000'
restart: unless-stopped
🔍 Troubleshooting
Common Issues
API Key Issues
Error: Unauthorized (401)
# Solution: Check your API key format and validity
Rate Limiting
Error: Too Many Requests (429)
# Solution: The server automatically handles rate limiting with exponential backoff
Connection Issues
Error: Network timeout
# Solution: Check your internet connection and Float API status
Debug Mode
Enable detailed logging:
LOG_LEVEL=debug
This will show all API requests/responses and internal operations.
🤝 Contributing
We welcome contributions! Please see our Contributing Guidelines for details.
Development Setup
# Clone the repository
git clone https://github.com/asachs01/float-mcp.git
cd float-mcp
# Install dependencies
npm install
# Set up environment
cp .env.example .env
# Edit .env with your Float API key
# Run in development mode
npm run dev
Code Quality
# Run linting
npm run lint
# Format code
npm run format
# Type checking
npm run typecheck
📄 Documentation
- API Documentation - Complete API reference
- Tool Reference - Detailed tool documentation
- Usage Guide - Usage patterns and examples
- Integration Testing - Testing guide
- Claude Integration - Claude Desktop setup
📊 Project Status
- ✅ Complete Float API Coverage - All Float v3 endpoints implemented
- ✅ Production Ready - Full error handling, rate limiting, and logging
- ✅ Well Tested - Comprehensive integration test suite
- ✅ Type Safe - Full TypeScript with Zod validation
- ✅ Docker Ready - Production-ready containerization
- ✅ MCP Compatible - Full Model Context Protocol compliance
🗺️ Roadmap
- [ ] Float API v4 Support - When released by Float
- [ ] Enhanced Caching - Optional Redis caching layer
- [ ] Webhook Support - Real-time Float event notifications
- [ ] Bulk Operations - Enhanced bulk import/export tools
- [ ] Custom Reports - Advanced reporting and analytics
- [ ] Multi-tenant Support - Multiple Float account support
📝 License
This project is licensed under the MIT License - see the LICENSE file for details.
🙋 Support
- 📖 Documentation: See the docs directory
- 🐛 Bug Reports: GitHub Issues
- 💬 Discussions: GitHub Discussions
🏷️ Tags
float mcp model-context-protocol project-management resource-management time-tracking typescript api-integration claude ai-tools
Built with ❤️ for the Float and MCP communities
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。