Terms our team uses internally

Every piece of research ends in one of these two words. ACT links to the work it changed. REJECT gives a one-line reason.

The page where every public post waits for a yes from Robb, or from the brand's manager where Robb has handed that over. Nothing posts before that yes.

A small watcher that raises an alarm when something that should have happened didn't, like a lead email that never went out.

One piece of work on a team board. Every card has an owner, a count and a check. Once it moves to To do, it gets a due date. Cards live on a board in Basecamp, the tool we use to track work.

The columns a card moves through on its Basecamp board: Ideas, Not now, Roadmap, To do, In progress, Blocked and Done. Only Robb moves a card to Roadmap. Ideas have no date and are ranked. Everything in To do has a date.

The agent that keeps every desk, server and safety check running.

The agent that runs Robb's calendar, inbox and queue, and chases anything running late.

Approving a type of piece once. After that, each new piece of that type ships when it passes its quality review. That one approval counts as the yes on the approval board.

Every brief says how many ("5 posts") and how we'll prove the work is done ("5 posts waiting on the approval board").

The most urgent fixes. They can't wait.

One AI agent with one job, like the content desk or the design desk. Each desk has a name and a title.

The instruction sheet each desk reads when it starts. A program builds it, and nobody edits it by hand.

Work is done when it works for a person on the live page or in the inbox. Something other than the maker has to check it.

A date on a card is a promise. Work we can finish today is due today.

A fleet is a team of AI agents, each with its own job, working together in one business, following shared rules, with a person approving what matters.

A Basecamp to-do list of problems in how our system works. Robb and the chief engineer decide them Mondays and Thursdays. The goal is to fix what broke in the system. Nobody gets blamed.

An agent saves its open work and notes to memory. Then its session can be cleared without losing anything. It's also Robb's one-word command to a desk: say 'flush,' and it saves and clears.

One last read that asks, "What would one more pass change?" Then we make that change or ship the piece.

A check that work must pass before it moves on. If a gate is wrong, we fix the gate. We don't step around it.

A rule check that runs before an agent acts. It blocks the action if it breaks a rule, like posting as Robb without his yes.

A short note of what Robb asked for, agreed on before any building starts.

The tests a problem must pass to go on the problem log. It involved an outside tool, we have the exact error, it cost us time, and the fix was checked.

No code change joins the main copy until a second AI reviewer has read it and every serious finding is fixed and its tests pass.

See it in the log: AI Agent Removed a GitHub PR Hold Label: Check Every Commit

Code added to the main version isn't done until the live page shows it.

Every night at 1 AM, each desk saves its work and restarts, so it starts the day on the newest rules.

The rules every desk follows: how to talk to Robb, who decides what, and what counts as done.

Giving a task a later due date than the work needs. Up to 15 minutes late, with a heads-up first, isn't a miss.

A private message between Robb and one desk. It's a Basecamp feature.

Our public list of problems we hit, each with the exact error and the fix that worked. It's at bluhook.com/log.

See it in the log: bluhook.com/log

A separate AI that did not make the piece scores it out of 10 before a person sees it. A low score, or any problem it marks as blocking, sends the piece back.

Proof that an agent did what it says. One example: how many frames of a video it looked at, out of the total, before it says "I watched the video."

See it in the log: AI agent says it watched a video but didn't? Make it prove every frame

A reviewer whose job is to break the work. It looks for the hole before someone outside finds it.

Any desk can ask another desk for work by putting a card on that desk's Basecamp board. The other desk accepts or offers a different plan within 4 hours. If Robb says now, the other desk answers within 15 minutes.

A picture puzzle or code a website shows to prove a person is there. Agents never solve them. A person does.

A saved set of steps an agent loads when a task needs it, so the method is the same every time.

A private copy of a website. We check changes there before they go live.

The desk that checks every rule change before it goes live. It asks: does this clash with another rule, and which desks need it?

The GM, the chief engineer and the chief of staff. They clear blockers together. They can also say yes to a new key or login that costs nothing. Anything that costs money is Robb's call.

A disagreement on a card gets two rounds at most. After that, one question goes to Robb with both views.

Each desk's daily count of finished work, shown next to its target. The target is the least we'll do, not the most.

A short-lived helper agent that a desk sends to do one job. When the job is done, the worker is gone.