Beyond data managers: prompt builders

For years we trained people to be good data managers — curators of inputs, owners of spreadsheets, keepers of context. That skill still matters. But the job description is expanding.

The new seat: prompt builder

A prompt builder does not write clever one-liners. They structure work so a model — or an agent — can execute reliably. They think in CRIT:

• Context — what situation are we in?
• Role — who is the model playing?
• Instructions — what must happen, in what order?
• Target — what does “done” look like?

Vague asks produce vague outputs. CRIT turns a hallway conversation into something you can hand to a colleague — or save as a skill.

From prompts to skills

The payoff is reuse. A one-off prompt dies in a chat thread. A CRIT-structured prompt becomes a template, then a skill, then part of how your team runs morning briefings, customer prep, or incident summaries.

At CIT we see the pattern everywhere: the teams that win are not chasing the newest model. They are building a library of prompts that encode how we work here.

What to do Monday

Pick one recurring task — a weekly status, a prospect brief, a post-incident recap. Write it in CRIT. Run it twice. Fix what breaks. Name it. Share it in your team forum.

That is how prompting scales: not hero prompts from one power user, but a growing set of skills anyone can run.