Taeyang Kim

I ve been thinking about a product-design problem that seems simple at first, but gets complicated very quickly:

When should an AI agent just do the thing and when should it stop and ask for confirmation?

The more actions an agent can take, the more useful it becomes.

But the more autonomy you give it, the more important trust becomes.

We spent three months building an open-source, local-first AI work journal that captures daily activity and turns it into clear timelines, worklogs, and updates. Yesterday, the Product Hunt community helped it reach the top position.

Thank you to everyone who supported the launch, shared their experience, left a comment, or recommended Meridian to someone else. The launch is over, but the work continues. We are reviewing what people shared and improving the onboarding, activity timeline, and drafted updates.

If you have tried Meridian, we would value hearing honestly what worked well and what still feels unclear.
GitHub: https://github.com/Meridiona/mer...

Product Hunt: https://www.producthunt.com/prod...

One of the teams using MonoCloud made a decision most founders are too nervous to make: they turned off passwords completely.

no fallback, no 'or create a password instead.' passkeys by default, magic link as the backup, done.

here's what actually changed after they did it.