THE NOVA AI WEEKLY
Operator systems, frameworks, and deployable AI workflows for corporate professionals.
Welcome to Issue #2 of The Nova AI Weekly.
If you’re here from launch week — thank you. The response was better than we expected.
The most common message I received: “I never thought about structuring my prompts this way.”
That’s exactly the gap Nova AI exists to close. Today we’re going deep on the system itself.
THE FRAMEWORK
Most people who use AI every day still treat it like a search bar.
They type a question. They read the output. They’re mildly disappointed. They move on.
The problem isn’t the model. It’s the input structure — or the complete absence of one.
The RCTF Framework is four inputs. Any AI model. Professional-grade output, every time.
R — Role
Give AI an identity before you give it a task.
“You are a…” is the most powerful two words in prompting — not because it’s magic, but because it calibrates everything that follows. Vocabulary, depth, tone, perspective, the assumptions the model makes about what you need.
An AI told “you are a senior McKinsey consultant” does not produce the same output as an AI told “summarize this.” Same prompt length. Completely different result.
Role is not a job title. It’s a lens. Assign the right lens before you assign the task.
C — Context
What AI doesn’t know, it fills in wrong.
Most prompts are context-free. The model has no idea who the audience is, what the stakes are, what has already been tried, or what a good outcome looks like. So it generates a statistically average response — which is, by definition, generic.
Context is the background information that transforms a generic answer into a precise one. One to three sentences. Answer these four questions before you write the task:
Who is the audience? What do they care about? What has already happened? What does “done” look like?
That’s it. Four questions. Every output improves.
T — Task
Specify what “done” looks like — not just what to do.
Most people give AI a verb. Summarize. Draft. Analyze. That’s a direction, not a specification.
An operator specifies the output: “Write a 3-paragraph executive summary that leads with the key finding, supports it with two data points, and ends with the one recommended action.” That’s a task. The verb is just the beginning.
The more precisely you define done, the less editing you do after.
F — Format
Tell AI exactly how to package the output.
Length. Structure. Tone. Header style. Whether to use bullets or prose. Whether to include a subject line. Whether to write in first person or third.
Format determines whether an output is usable or whether it has to be rewritten. Most prompts leave this completely unspecified — and then the user wonders why they always have to clean up the output.
You’re not editing AI. You’re compensating for incomplete instructions. Add the format specification and the editing disappears.
RCTF IN PRACTICE — Two Versions of the Same Prompt
The version most people write:
“Write a follow-up email to my client.”
Result: three paragraphs of generic, professional-sounding text that could apply to any industry, any client, any situation.
The operator version:
“You are a senior account manager at a consulting firm. My client Sarah is VP of Operations at a mid-size manufacturer. We had a discovery call 48 hours ago — she expressed interest but asked for a pricing breakdown before her CFO conversation, which I’m not invited to. Write a follow-up email that acknowledges the brief delay, references her CFO approval process as the next step, and gives her one sentence she can use to introduce the pricing discussion internally. Professional but warm. Under 120 words. Subject line included.”
Same AI. Same model. Completely different output.
One of those emails gets sent directly. The other gets rewritten.
THIS WEEK’S OPERATOR ASSIGNMENT
Pick one email you need to write this week.
Before you open Copilot, ChatGPT, or Claude — write out these four things first on paper or in a notes app:
Role: Who should AI be acting as?
2. Context: What’s the situation, the audience, the stakes?
3. Task: What does the finished output look like — specifically?
4. Format: Length, tone, structure?
Then paste all four into your AI tool as the prompt.
Notice the difference in the output.
That’s RCTF. Four inputs. One week of practice. The methodology sticks.
FROM THE OPERATOR PACK
The Nova AI Operator Pack includes 21 pre-built RCTF prompts across 5 professional categories:
Email — Follow-ups, de-escalations, executive updates, cold outreach
Meetings — Pre-meeting briefs, post-meeting action extraction, stakeholder prep
Research — Document compression, competitive intelligence, decision briefs
Writing — Executive summaries, policy drafts, performance review narratives
Decisions — Option analysis, risk assessment, recommendation framing
Each prompt is deployment-ready. Fill in the brackets, paste into your AI tool, use the output.
Get the Operator Pack — $97: https://novamedia42.gumroad.com/l/ozmzce
NEXT ISSUE
We’re going deep on one specific RCTF workflow: the Meeting Prep System.
Before your next important meeting, you’ll have a prep brief with 3 agenda items, 2 questions to ask, 2 objections to prepare for, and one thing not to bring up — generated in 90 seconds.
It publishes next Tuesday.
Forward this to one person at work who uses AI every day and feels like they’re still not getting what they should out of it. That’s who this is for.s