Clockify MCP Server
Enables time tracking and project management through the Clockify API. Supports starting/stopping timers, logging time entries, managing projects and tasks, and generating reports with natural language commands.
README
Clockify MCP Server
An MCP (Model Context Protocol) server that provides tools for interacting with the Clockify time tracking API.
Installation
npm install -g @yikizi/clockify-mcp
Or use directly with npx:
npx @yikizi/clockify-mcp
Features
- Workspace Management: List workspaces, get current user, list workspace users
- Project Management: List/create projects, list/create tasks
- Time Tracking: Start/stop timers, log time entries, update/delete entries
- Reporting: Generate summary and detailed reports with grouping and filtering
- Tags & Clients: List and create tags, list clients
Setup
1. Get your Clockify API Key
- Go to Clockify User Settings
- Scroll to "API" section
- Click "Generate" to create an API key
2. Configure Claude Code
Add to your Claude Code MCP settings (~/.claude.json or via Claude Code settings):
Option A: CLI argument (recommended)
{
"mcpServers": {
"clockify": {
"command": "npx",
"args": ["@yikizi/clockify-mcp", "--api-key", "your-api-key-here"]
}
}
}
Option B: Environment variable
{
"mcpServers": {
"clockify": {
"command": "npx",
"args": ["@yikizi/clockify-mcp"],
"env": {
"CLOCKIFY_API_KEY": "your-api-key-here"
}
}
}
}
Option C: .env file
Create a .env file in your working directory:
CLOCKIFY_API_KEY=your-api-key-here
Available Tools
Convenience Tools (Recommended)
These auto-detect your user and workspace:
start_timer- Start a timer (just provide description and optional project)stop_current_timer- Stop whatever timer is runningget_running_timer- Check if a timer is running and see elapsed timelog_time- Log completed time with human-readable duration ("1h30m", "2h", "45m")
Workspace Tools
get_workspaces- List all workspacesget_current_user- Get authenticated user infoget_workspace_users- List users in a workspace
Project Tools
get_projects- List projects (with archive filter)create_project- Create a new projectget_project_tasks- List tasks in a projectcreate_task- Create a task
Time Entry Tools (Low-level)
get_time_entries- Get time entries for a user (with date/project filters)create_time_entry- Create a time entry (usestart_timerorlog_timeinstead)stop_timer- Stop timer for specific user (usestop_current_timerinstead)update_time_entry- Update an existing entrydelete_time_entry- Delete a time entry
Reporting Tools
get_summary_report- Aggregated report grouped by project/user/clientget_detailed_report- Detailed list of time entries with filters
Tag & Client Tools
get_tags- List all tagscreate_tag- Create a new tagget_clients- List all clients
Usage Examples
Once configured, you can ask Claude things like:
- "Start a timer for working on the API project"
- "Stop my timer"
- "Log 2 hours for documentation work"
- "Is my timer running?"
- "Generate a summary report for last week"
- "How much time did I log this month?"
- "List all my projects"
Changelog
v1.2.2
- Config: API key can now be passed via
--api-keyCLI argument - Config: API key can be read from
.envfile in current directory - Improvement: Better error message showing all config options
v1.2.1
- Bugfix:
get_running_timernow properly finds running timers (was only checking 1 entry) - Bugfix:
stop_current_timernow correctly stops timers - Improvement:
stop_current_timeraccepts optionalendtime for backdating
v1.2.0
- Human-readable durations in responses (
1h 30minstead ofPT1H30M) - Relative period filters for
get_time_entries:today,yesterday,this_week,last_week,this_month,last_month get_time_entriesnow defaults to current user/workspace (no IDs needed!)- Responses include total time across all entries
v1.1.0
- Added convenience tools:
start_timer,stop_current_timer,get_running_timer,log_time - Auto-detect user and workspace (no need to call
get_current_userfirst) - Human-readable duration format ("1h30m") for
log_time - Input validation to catch date typos early
- Warns if timer is already running when starting a new one
v1.0.0
- Initial release with core Clockify API support
License
MIT
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。