super-productivity-mcp
Enables interaction with Super Productivity's local REST API to manage tasks, projects, and time tracking through natural language.
README
super-productivity-mcp
An MCP server and two Claude skills for Super Productivity's Local REST API, plus a one-line installer so a teammate can onboard by pointing their Claude co-worker at this repo.
Super Productivity's API runs on http://127.0.0.1:3876, requires no auth
(localhost-only), and must be turned on manually via Settings → Misc →
Enable local REST API.
The two skills
super-productivity-onboarding— drives setup itself: detects the user's OS, briefs them on what it's about to install and why, gets explicit go-ahead, then installs Super Productivity (if missing), Bun, builds and registers the MCP server, installs both skills, and verifies everything works.super-productivity-daily-schedule— day-to-day use: planning the day, adding/checking off tasks, time tracking, end-of-day review, via thesp_*MCP tools.
Onboard someone else
Point their Claude Code at the raw URL of ONBOARDING.md
and ask it to follow the steps — it directs Claude to read and execute
skill/onboarding/SKILL.md, which will:
- Detect their OS and brief them on the plan, then wait for a clear go-ahead before installing anything.
- Check for the Super Productivity desktop app (and open the download page if it's missing).
- Install Bun if needed (no admin rights required).
- Download this repo and compile the MCP server into a single standalone executable — no Node/Bun required at runtime after this.
- Install both Claude skills into
~/.claude/skills/. - Register the server with Claude Code:
claude mcp add -s user super-productivity -- <path-to-exe>. - Ask the user to flip the "Enable local REST API" checkbox in the app's
Settings — the one step that can't be scripted safely, since it's
in-app UI state rather than a config file — then verify with
sp_health.
Manual equivalent (Windows PowerShell):
irm https://raw.githubusercontent.com/kailasas-auspicious/super-productivity-mcp/main/install/install.ps1 | iex
Manual equivalent (macOS/Linux):
curl -fsSL https://raw.githubusercontent.com/kailasas-auspicious/super-productivity-mcp/main/install/install.sh | bash
What's in here
src/— the MCP server (TypeScript), wraps every endpoint of the Local REST API (sp_list_tasks,sp_create_task,sp_update_task,sp_start_task,sp_get_status,sp_list_projects, etc).skill/onboarding/SKILL.md— drives first-time setup (OS detection, briefing, install, verify).skill/daily-schedule/SKILL.md— day-to-day task/schedule management via thesp_*tools.install/— the underlying install scripts the onboarding skill runs.ONBOARDING.md— a thin pointer for Claude, not a human, to the onboarding skill.
Local dev
bun install
bun run dev # runs src/index.ts directly via tsx
bun run build # tsc -> dist/
bun build ./src/index.ts --compile --outfile dist/super-productivity-mcp
Point Claude Code at your local build for testing:
claude mcp add -s user super-productivity -- bun run --cwd /path/to/super-productivity-mcp src/index.ts
Config
Set SP_API_URL if Super Productivity's API isn't on the default
http://127.0.0.1:3876.
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。