I've been building too many things at once. Today I drew a line.

Three projects going forward — and only three:

• gl-elevatedaily — the first AI app where there's a real path to income. Manifestation system, daily brief, oracle layer.
• gl-options — a personal options-selling tool for TastyTrade. I'm building it for me first. If it earns its keep in my morning routine, it productizes.
• gl-neil — this site. The build-in-public surface. Letter, journal, blog, chat. The professional image, in long form.

Everything else — gl-defi, gl-academy, the half-built infra projects — paused. Not killed. The work doesn't disappear; the attention does.

The pattern I keep falling into: a new idea looks adjacent enough to current work that it feels like leverage to start it. A week in, it's its own project, with its own scaffold, its own decisions, its own gravity. Three of those running concurrently is a recipe for none of them shipping.

So: narrowing. Not because the other things are bad — most of them are interesting. Because optionality without delivery is just sophisticated procrastination.

What changes day-to-day:

• Mornings start in one of the three. Not "what should I work on?" — "which of the three?"
• Research goes into the three. Reading goes into the three. Side-quests get a one-line note in the journal and stay there.
• When something genuinely matters that isn't in the three, I'll know — because it'll keep coming back. Until then, the rule holds.

Building in public means showing the constraint, not just the wins. This is the constraint.