Tycana
AI-native productivity backend that gives your AI assistant persistent memory, pattern awareness, and computed intelligence about your work. 14 MCP tools for task management, daily planning, weekly review, and personal context.
README
<p align="center"> <img src="tycana-logo.svg" alt="Tycana" width="300"> </p>
<p align="center"> <strong>AI-native productivity backend — your AI thinks, Tycana remembers.</strong> </p>
<p align="center"> <a href="https://www.tycana.com/getting-started/">Get Started</a> · <a href="https://www.tycana.com">Website</a> · <a href="https://www.tycana.com/mcp/">MCP Docs</a> </p>
What is Tycana?
Your AI assistant is brilliant but amnesiac. Every conversation starts from zero — no memory of what you captured yesterday, no awareness of what's overdue, no sense of your work patterns. You re-explain your priorities every morning. Tycana fixes this.
Tycana is a productivity backend built for AI reasoning, not human browsing. Connect your AI once via MCP (Model Context Protocol), and every conversation automatically includes your full work picture: active projects, upcoming deadlines, blocked items, and computed intelligence about your patterns.
This isn't another task app with an AI bolted on. There's no UI to check, no lists to reorganize, no weekly reviews to dread. Tycana is infrastructure that sits behind your AI conversation, making every interaction smarter.
Quick Start
1. Sign up
Start a 14-day free trial — no credit card required.
2. Connect your AI
Add Tycana to your MCP client configuration:
{
"mcpServers": {
"tycana": {
"type": "streamable-http",
"url": "https://app.tycana.com/mcp",
"headers": {
"authorization": "Bearer YOUR_TOKEN"
}
}
}
}
Most MCP clients (Claude Desktop, Claude Code, Cursor) support OAuth — just connect and authorize when prompted. The bearer token config above is for clients that don't support OAuth flows.
3. Start talking
Just use your AI normally. Say things like:
- "Remind me to update the deployment docs before Friday"
- "Plan my day"
- "How's Project X going?"
- "What should I work on next?"
Tools
Tycana provides 14 MCP tools:
Task Management
| Tool | Description |
|---|---|
capture |
Create a new task, note, or idea with optional project, priority, due date, and effort estimate |
complete |
Mark an item as done with an outcome (resolved, wont_do, delegated, deferred) |
update_item |
Update any field on an existing item — title, body, priority, due date, project, effort |
delete_item |
Permanently delete an item |
list_items |
List items filtered by status (active, completed, all), project, or priority |
get_item |
Retrieve full details for a specific item by ID |
search |
Full-text search across all items |
Intelligence & Planning
| Tool | Description |
|---|---|
plan_day |
Generate a prioritized daily plan based on available hours, deadlines, and your work patterns |
what_next |
Get a smart recommendation for what to work on based on available time (quick/hour/deep) |
review |
Weekly review — surfaces overdue items, stalled work, slip patterns, and velocity trends |
get_context |
Pull full context for a scope (today, project, everything) to ground AI responses |
Organization & Memory
| Tool | Description |
|---|---|
relate_items |
Create relationships between items (blocks, parent, related) |
remember |
Store personal context — preferences, work style, facts your AI should always know |
cleanup_project |
Archive or bulk-complete items in a finished project |
Key Features
- Persistent memory — Your AI remembers everything across conversations
- Computed intelligence — Velocity tracking, slip rate detection, stale work alerts, effort calibration
- Confidence-gated signals — Intelligence only surfaces when backed by enough data
- Personal context — Learns your preferences, work style, and priorities over time
- Calendar feed — Tasks appear alongside your meetings with effort-based lead times
- Daily email digest — Smart-scored briefing of what needs attention
- Project relationships — Items connect, block each other, have dependencies
- Works everywhere — Claude Code, Claude.ai, ChatGPT, Cursor, any MCP-compatible client
How It Works
You <-> Your AI Assistant <-> MCP <-> Tycana Backend
|
PostgreSQL
(your data)
Tycana is a hosted service. Your AI connects to https://app.tycana.com/mcp via streamable HTTP with OAuth 2.1 authentication. All data is stored securely and belongs to you — full export and deletion available at any time.
Pricing
One plan. Everything included.
- $15/month or $150/year (2 months free)
- 14-day free trial, no credit card required
- No feature tiers, no per-seat pricing, no upsells
Compatibility
Tycana works with any MCP-compatible client:
- Claude Desktop
- Claude Code (CLI)
- Claude.ai
- ChatGPT (via MCP)
- Cursor
- Windsurf
- Any client supporting streamable HTTP transport
Links
License
This repository contains documentation and configuration for the Tycana MCP server. The Tycana service itself is proprietary software. See Terms of Service.
This documentation is released under 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。