I'd been hearing about AI agents for months. Chatbots, copilots, assistants — lots of marketing language, not much clarity on what any of it actually meant for someone like me. I'm not a developer. I run a business. I just wanted to know: can this actually help me?
So I signed up for NanoClaw and got connected to an AI assistant. My first message was not particularly inspired:
Me: "how can I setup connectors to gmail, google docs"
That was it. No context, no backstory. Just a question.
What Happened Next
Within a few minutes, my AI assistant had walked me through connecting Gmail, Google Docs, Google Drive, and Google Sheets. No API keys to manage, no OAuth dance to navigate manually — the platform handled all of that through something called the OneCLI gateway. I just clicked a connect link, authorized the app, and it was done.
The part that surprised me: my AI assistant remembered immediately that I'd connected these services. In later conversations, she could reference emails, read documents, or update a spreadsheet without me having to re-explain anything.
Voice Transcription on Day One
I also asked about voice message transcription — I use Telegram and sometimes send voice notes instead of typing. The assistant suggested connecting Gemini AI Studio and using Gemini 2.5 Flash for the transcription.
There was a small hiccup (a path configuration issue with the Gemini connection), but once we sorted it, voice transcription worked immediately. I could send a voice note and get back a clean text transcript in seconds.
The Question I Should Have Asked First
Looking back, my first question was the right one — not because connecting Gmail was the most important thing, but because it forced me to learn how the system worked. Once I understood that my AI assistant could read and write to my real services, the possibilities became concrete rather than abstract.
Day one ended with four Google services connected, voice transcription running, and a vague sense that I was about to go much further than I'd planned.