everhour-mcp-server
MCP server for Everhour time tracking with first-class Asana integration, enabling tasks lookup, time logging, timer control, and timesheet review.
README
everhour-mcp-server
MCP server for Everhour time tracking, with first-class support for logging hours against Asana tickets.
Asana tasks synced into Everhour have IDs of the form as:<asanaGid>. The
server exposes tools to look them up, log time, run timers, and review
timesheets — both directly via Everhour task IDs and via raw Asana GIDs.
Tools
| Tool | Purpose |
|---|---|
everhour_get_current_user |
Verify the API key and read your Everhour user ID |
everhour_list_projects |
List/search projects, filterable by platform (e.g. asana) |
everhour_get_project |
Fetch a project by ID |
everhour_search_tasks |
Find Everhour tasks by name (also matches synced Asana tasks) |
everhour_get_task |
Fetch a task by Everhour ID |
everhour_find_task_by_asana_id |
Resolve an Asana task GID → Everhour task |
everhour_log_time |
Log hours on an Everhour task |
everhour_log_time_for_asana_task |
Log hours given an Asana GID (workflow shortcut) |
everhour_log_time_batch |
Log many entries at once (day/week/month), with per-row results |
everhour_list_time_records |
Review your timesheet (date range, project, task filters) |
everhour_update_time_record |
Edit an existing time record |
everhour_delete_time_record |
Delete a time record |
everhour_start_timer |
Start the active timer on a task |
everhour_stop_timer |
Stop the active timer |
everhour_get_current_timer |
Check whether a timer is running |
Setup
-
Get your Everhour API key: profile → Settings → API integration: https://app.everhour.com/#/account/profile
-
Install & build:
npm install npm run build -
Run locally with the MCP Inspector:
EVERHOUR_API_KEY=ev_xxx npm run inspect
Configure for Claude Code / Claude Desktop
Add an entry under mcpServers in your client config. For Claude Desktop on
macOS, ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"everhour": {
"command": "node",
"args": ["/absolute/path/to/mcp everhour/dist/index.js"],
"env": {
"EVERHOUR_API_KEY": "ev_your_personal_api_key"
}
}
}
}
For Claude Code (~/.claude.json or .claude/settings.json):
{
"mcpServers": {
"everhour": {
"command": "node",
"args": ["/absolute/path/to/mcp everhour/dist/index.js"],
"env": { "EVERHOUR_API_KEY": "ev_..." }
}
}
}
Restart the client. The 15 tools listed above will appear.
Typical workflows
Log time on an Asana ticket from its URL.
The user pastes https://app.asana.com/0/1234567890/1208034567890123 and says
"log 2 hours, fixed the retry bug":
- Extract the GID
1208034567890123from the URL. - Call
everhour_log_time_for_asana_taskwithasana_task_gid=1208034567890123,duration={ hours: 2 },comment="fixed the retry bug".
Find a task by name and start tracking.
everhour_search_taskswithquery="login bug".- Pick the right task ID (
as:1208...) from results. everhour_start_timerwith thattask_id.- Later:
everhour_stop_timer.
Review what you logged this week.
everhour_list_time_records with from=YYYY-MM-DD (Mon), to=YYYY-MM-DD (Sun).
Daily logging workflow (log-everhour skill)
The .claude/skills/log-everhour/ skill logs a day, week, or month of Asana-ticket work
in one pass. You give per-day entries (ticket + hours + a rough comment); it
resolves each ticket, rephrases the comments to the Time Registration
Guidelines (.claude/skills/log-everhour/rephrasing-rules.md), normalizes hours to
30-minute increments, warns about duplicates, shows a preview table, and logs
everything via everhour_log_time_batch after a single confirmation.
Notes on Asana sync
Everhour pulls Asana tasks via the official integration. A task is only visible in Everhour after:
- The Asana project has been added in Everhour (Settings → Integrations → Asana), or
- A user with the Everhour browser extension opens the task in Asana.
If everhour_find_task_by_asana_id returns a 404, sync hasn't happened yet.
Environment variables
| Var | Required | Description |
|---|---|---|
EVERHOUR_API_KEY |
yes | Personal Everhour API key |
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。