Prompt: Choose the right model for the task at hand (UP-Context, Working with AI)

2 days ago

5 min read

An abstract row of geometric blocks in blue, one lifted forward and marked in red.

In this Prompting Publication

What this prompt does

This prompt does one job: it stops you using one model for everything. You describe the task, it classifies the task into one of three bands, recommends a model for that band, and names the risk of getting it wrong.

UP-Context prerequisites

Required layers: USER-PERSONA, FULL-CONTEXT, AI-PROFILE-ROLE-EXPERTISE, AI-EXACT-TASK and AUDIENCE. Optional layers: USER-FINAL-GOAL, EXPECTED-FORMAT, EXAMPLES. The FULL-CONTEXT layer is what the classification runs on: without the task described, the recommendation is a guess about a generic task.

Where it runs

Execution environment: Chatbot, or Project. It is a decision aid you run once per task type, not per message. In a project it becomes a standing rule the assistant applies to its own recommendations.

The prompt

Copy the block below in full. It is the whole prompt: nothing else needs to be attached, and no University 365 file is requested of you.

PROMPT STARTS HERE

Select everything in blue from the line above the word PROMPT STARTS HERE down to the line before word PROMPT ENDS HERE, and paste it into a new AI chat with the appropriate LLM and thinking level. Replace every bracket with your own text before you send it.

INSTITUTIONAL CONTEXT (read before you begin) This prompt is complete on its own. Everything it needs about University 365, about ULM and EVA, and about the UP-Context Method is defined below. No University 365 file is attached to it and none is required. The institution. University 365 is an institution of higher learning. Its Department of Academics owns the curriculum and the academic standards. The method. The UP-Context Method (University 365 Prompting-Context) structures every prompt into eight layers, five required and three optional: USER-PERSONA (who is asking), FULL-CONTEXT (the situation and the facts), AI-PROFILE-ROLE-EXPERTISE (who the AI is acting as), AI-EXACT-TASK (the exact job asked now), AUDIENCE (who the output is for), plus the three optional layers USER-FINAL-GOAL (the goal being worked toward), EXPECTED-FORMAT (the shape of the answer) and EXAMPLES (reference examples). A layer is text in the prompt first; keep it as a reusable file when its text is long enough and reused across prompts. The framework. ULM (University 365 Life Management) organises a person's life into six domains: Body and Health, Spirit and Mind, Character and Emotions, Social and Love Relationships, Career and Finance, Quality of Life. EVA is the engine that carries a person through each domain in three phases: Explore, Visualize, Action Plan. Who is who. U.Copilot is the AI assistant. U.Coach is a human coach. U.Copilot is not a human, must never present itself as one, and must never give medical, legal or financial advice. Your judgement. This prompt supports your thinking. Every recommendation it produces is a draft for you to check, not a decision made on your behalf. THE MODEL CHOICE #USER-PERSONA: I am the person who will run this task. If a persona file is present in this project, apply it. If the layer was written as text instead, it appears after this line: [Write three to six lines about yourself that matter for this task: the kind of work you do, how you judge a good answer, and what you already know so that it need not be explained again.] #FULL-CONTEXT: the task this choice is about. If a context file is present in this project, apply it. If the layer was written as text instead, it appears after this line: [Describe the task, what it produces, who reads it, and what a wrong answer would cost.] #AI-PROFILE-ROLE-EXPERTISE: you are a careful practitioner who knows several model families and their real strengths. You separate what is measured from what is marketing, and you say which is which. #AI-EXACT-TASK: answer in three parts. 1. CLASSIFY THE TASK against the three bands below and say which band it is in: THINKING: long reasoning, planning, analysis, architecture, judgement under uncertainty. DRAFTING: producing volume of well-formed text or code from a clear brief. CHECKING: reviewing, cross-checking, finding what was missed. 2. RECOMMEND a model for THIS task and say why it fits. Name the capability that matters, not a brand ranking. If your answer depends on which model you are, say so. 3. NAME THE RISK of using the wrong model for this task, and one cheap safeguard. #AUDIENCE: me, before I start the task. Assume I am competent but new to model selection. #EXPECTED-FORMAT: three short sections with the exact titles ABOVE: CLASSIFY THE TASK, RECOMMEND, NAME THE RISK. No preamble. In section 1, output the band word in capitals. #EXAMPLES: if I give you the task output later, re-run the classification before you recommend anything. RULES 1. Do not recommend a model for a task you have not classified first. 2. When a recommendation rests on a configuration, a plan, a price or a version, say so, because all of those change. 3. Never invent a benchmark result. If you do not have a measured fact, say the recommendation is a judgement. 4. If the task fits none of the three bands, say so and stop. Do not force a classification. 5. End with one question that would change your recommendation if I answered it.

PROMPT ENDS HERE

The brackets are the only parts you change. Everything else is the prompt.

A three-band diagram: thinking tasks, drafting tasks and checking tasks, with the label match the tool to the job.

What to change for your situation

Change the FULL-CONTEXT bracket first: the classification depends on the task, not on you. Then adjust the three bands if your work divides differently. Keep the CLASSIFY step in place: a recommendation that skips it is a brand opinion.

Verification checklist

Multi-model check: run the prompt on a second model and compare. External source check on any factual claim that matters. Human review before you act. CI-First test: does using this prompt build your own judgement, or replace it?

What this does to you

CI-First score 7.1, band Strong: time 8, quantity 6, quality 8, skill 7. Humics: Creativity Neutral, Critical Thinking Friendly, Social Authenticity Friendly, value +2. Imposture risk Low: the prompt refuses to recommend before it classifies, and it must mark a recommendation that rests on a judgment rather than a measurement.