How the agent runs it

This system is built to be operated by an AI agent with you as the decision-maker. The agent does the reading, filing, chasing, and sweeping. You make calls. These are the loops that make it work.

Morning briefing

Every morning the agent reads what changed overnight - new captures, new email that matters, calendar for the day - and gives you one short briefing. You start the day knowing the state of the board without opening anything.

Daily blocked-on-you digest

Anything in the system that is waiting on YOUR decision gets collected into one digest. Blocked means "needs you" - nothing sits in Blocked for any other reason. One list, once a day, everything that's yours to move.

Entertainment pass-through

Send the agent a link - a podcast, a YouTube video, a post - and it lands in Second Brain tagged by type (entertainment / AI to work with / potential revenue) and medium (Podcast, YouTube, X / Twitter). No questions asked, no friction. Entertainment passes through silently; anything with a business angle gets flagged.

Hourly sweep

Once an hour the agent sweeps: new rows it didn't create, calendar hygiene, drift between systems. Quiet when there's nothing. Reports only what matters.

The rules that keep it clean

Kanban conventions

Five statuses: Not started, In progress, Blocked, Waiting, Done. Blocked = needs your decision; Waiting = needs a third party. Every card gets Tags and a 1-10 Score. Revenue, health, and family get worked first. When a card is blocked on something specific, that something becomes its own sub-card under "Blocked On" - so the blocker itself is trackable work, not a note.

Activity Log conventions

One card per piece of work, opened when the work starts. Status Done plus an End date when it finishes. Where records how the work happened (Phone, Desktop, Laptop, Email, IRL) - over time it shows you where your work actually gets done. Value, Cost, Urgency, Confidence scored High / Medium / Low when a decision needs ranking; don't force-score junk.