Who This Is For

• macOS users who want their AI to know their schedule without manual copy-paste
• People who manage calendars via .ics files and want natural language access to them
• Anyone wanting schedule-aware LLM responses (free slot detection, conflict surfacing) without cloud sync
• Users who rely on LM Studio as a daily assistant and want context carried into every session

Installation

Load the built plugin folder in LM Studio.

Requirements

• macOS — Calendar.app source requires macOS with Calendar.app and Accessibility permissions granted
• Any platform — ICS file source works cross-platform

To grant Accessibility permission: System Settings → Privacy & Security → Accessibility → allow LM Studio (or Terminal if running lms dev).

Configuration

Tool: calendar

Single tool with four actions.

today

All events for today (or a specific date).

upcoming

Events in the next N days.

search

Find events matching a keyword in title or location.

free_slots

Find free blocks of at least N minutes on a given day.

All actions share these parameters:

• date — ISO date string YYYY-MM-DD (default: today)
• days_ahead — days to look ahead (upcoming only)
• keyword — search term (search only)
• min_gap_minutes — minimum free block size in minutes (free_slots only, default 30)

How Prompt Injection Works

When Inject Today into Prompt is enabled, the first message of every session gets prepended with today's events:

The LLM uses this to answer scheduling questions, suggest meeting times, or surface conflicts without you having to ask explicitly.

Example Workflows

Morning briefing:

"What does my day look like?" → calendar(action="today")

Find time for a meeting:

"When am I free for an hour tomorrow?" → calendar(action="free_slots", date="2025-06-11", min_gap_minutes=60)

Upcoming travel:

"Do I have anything scheduled next week?" → calendar(action="upcoming", days_ahead=7)

Find a specific event:

"When is my dentist appointment?" → calendar(action="search", keyword="dentist")