Why does my AI need a tool just to know the date and time?

Most AI assistants don't have access to your machine's clock and either guess or default to UTC, which causes wrong dates and off-by-hours scheduling. get_datetime returns your Mac's real current time, timezone and UTC offset so every time-relative task is accurate.

Does get_datetime send any data to the cloud?

No. It reads the current time from the Mac where LMCP runs and returns it directly to your AI client. Nothing is stored on a server and no API keys are required — everything runs locally on your device.

Which timezone does it report?

It reports the local timezone configured on the Mac running LMCP, including the timezone name (like America/New_York) and the current UTC offset, with daylight saving time already applied.

Get Current Date & Time on Your Mac with AI (LMCP)

The get_datetime tool lets your AI assistant read the current date and time directly from the Mac where LMCP runs — including the local timezone name and the UTC offset. This sounds trivial, but most AI assistants have no idea what time it is on your machine. They guess, hallucinate dates, or default to UTC. With LMCP, the answer is real. Just type something like "What's the exact date and time right now, and what timezone am I in?" and the AI returns your machine's actual clock, for example 2026-06-10 14:32 (America/Argentina/Buenos_Aires, UTC-03:00).

LMCP is a free, native macOS MCP server that gives AI assistants access to your real Mac apps and system locally. Download LMCP and your AI gains 150+ tools, this one included.

Which AI agents work?

Any MCP-capable assistant can call get_datetime. Desktop clients auto-configure over a local stdio connection, while web AIs connect through the LMCP Cloud Relay connector.

Desktop (auto-configure via local stdio): Claude Desktop, Cursor, VS Code (GitHub Copilot), Windsurf and Zed.
Web AIs (via the LMCP Cloud Relay connector): ChatGPT, Claude.ai (web), Grok and Perplexity.

Desktop clients pick up the tool the moment you restart them. Web AIs connect securely through the relay so they can reach your Mac without exposing it to the open internet.

Automation

On its own, knowing the time is useful — but the real power is chaining. Because get_datetime grounds the AI in the actual current moment, every time-relative request becomes accurate:

"Create a reminder for two hours from now" — the AI reads the clock, then calls the Reminders tool with the right absolute time.
"Schedule a calendar event for tomorrow at 9am my time" — no UTC drift, because the offset is known.
"Draft an email and note that it's end of day Friday here" — correct day and timezone in the body.
"Log this note with an accurate timestamp" — paired with Notes or Finder tools.

Without it, AI assistants routinely book meetings on the wrong day or off by several hours. get_datetime is the anchor that makes the rest of your scheduling tools trustworthy.

Context

This tool reads from the machine where LMCP runs — your actual Mac. That means the AI works with your real local time, your real timezone setting, and your real DST state, not a server-side approximation. When you ask the AI to coordinate across your calendar, messages and reminders, it shares the same sense of "now" that you do sitting in front of the screen.

Productivity

It is a small tool that removes a tax you pay constantly. Instead of correcting the AI ("no, it's Tuesday", "that's 3 hours off", "I'm in CET not PST"), it simply gets it right the first time. Over a week of scheduling emails, reminders and events, that is dozens of avoided back-and-forth corrections and zero mis-booked meetings. For anyone working across timezones, it quietly eliminates one of the most common sources of AI scheduling errors.

Privacy & GDPR

Everything runs locally on your Mac. The current time is read from your machine's system clock and returned straight to your AI client — no cloud copy, no logging on a server, nothing stored externally. LMCP needs no API keys and keeps your data on your device by design. Read more about why this is GDPR-compliant by architecture.