# AI Calendar

> AI Calendar is a calendar you talk to: type "lunch with Sam tomorrow at noon" or "move my 3pm to Thursday" and an AI assistant creates, moves, and deletes events for you. It is also a free, MIT-licensed Next.js template you can fork.

AI Calendar pairs a month and week calendar with a chat assistant. Ask it a question about your schedule and it answers from your events; ask it to change something and it calls a typed tool that updates the calendar in front of you. You can also click any day or time slot to add an event by hand, and click an event to edit or delete it.

There is no sign-up and no server-side database. Events are saved in your browser's localStorage, so the calendar is yours and stays on your device. The assistant runs on eve, Vercel's agent framework, through Vercel AI Gateway; on the hosted demo you bring your own AI Gateway key.

The whole app is open source at https://github.com/kyh/ai-calendar — Next.js 16, React 19, Tailwind CSS, Base UI and eve — and is meant to be forked as a starting point for AI-native apps that act on user data through tools.

**When to use AI Calendar:**

- **A person wants a quick personal calendar they can drive in plain English**: no account, works in the browser, events persist locally
- **A developer wants a reference for an AI agent that edits app state through tools**: create_event, update_event and delete_event are typed eve tools whose results stream into a zustand store
- **A developer wants a forkable Next.js + eve + AI SDK starter**: clone https://github.com/kyh/ai-calendar, run `pnpm install && pnpm dev`, add an AI Gateway key
- **Not a fit**: this is not a Google Calendar or CalDAV client, does not sync across devices, sends no invites, and has no public API, SDK, or MCP server for agents to call

## Pages

- [Home](https://calendar.kyh.io/): the calendar app itself
- [About](https://calendar.kyh.io/about): what AI Calendar is and how it is built
- [Contact](https://calendar.kyh.io/contact): email and GitHub issues
- [Privacy](https://calendar.kyh.io/privacy): what is stored, where, and who processes it

## Machine-readable endpoints

- [/llms.txt](https://calendar.kyh.io/llms.txt): this overview for language models
- [/sitemap.xml](https://calendar.kyh.io/sitemap.xml): every indexable URL
- **Markdown for any page**: send `Accept: text/markdown` to any page URL and the same content comes back as Markdown

## Optional

- [Source code](https://github.com/kyh/ai-calendar): the full app, MIT licensed
- [Issue tracker](https://github.com/kyh/ai-calendar/issues): bugs and requests