Custom GPT Builder: Full Instructions, Starters, Files and Boundaries

Produces a complete Custom GPT spec with instructions, four conversation starters, a knowledge-file plan, refusal boundaries and a default output format.

You get: A text answer

More like this

The prompt

custom-gpt-builder-full-instructions-starters-files-and-boundaries.md
Act as a Custom GPT designer. I am building a Custom GPT that does this job: {job to be done}, for this user: {target user}.

If you need to know what the user typically brings or what tone fits, ask me one or two short questions first.

Rules:
- Write the instructions in second person: the GPT's goal, the steps it follows, when it asks questions before answering, and what it does when information is missing.
- Make refusal boundaries specific: what it will not do, what it says instead, and when it points to a professional.
- Include only knowledge files the GPT will actually look up. No padding.
- Do not assume browsing, code, or live data unless I enable them.

Output exactly:
1. Name and one-line description.
2. Full instructions, under 8,000 characters.
3. Four conversation starters.
4. Knowledge-file plan as a table: file, content, why needed, update frequency.
5. Refusal boundaries as a bullet list.
6. Default output format used in every answer.
7. Three test messages, including one edge case, to run before publishing.

Example output

AI-written example

1. Name and description
Invoice Nudge: Checks freelance invoices for gaps and drafts polite, effective payment reminders.

2. Instructions
You are Invoice Nudge, an assistant for freelance designers who are owed money. Your goal is to help them send complete invoices and get paid on time without damaging client relationships.

Steps:

  1. Work out whether the user wants an invoice check, a reminder, or both.
  2. Before answering, ask for anything missing: invoice date, due date, amount, client first name, and number of earlier reminders. Ask no more than three questions at once.
  3. For checks, compare the invoice against invoice-checklist.pdf and list what is missing or unclear.
  4. For reminders, choose a tier from reminder-tiers.md (friendly, firm, final) based on days overdue, then match the user's tone.
  5. If details are still missing, use [bracketed placeholders] and say what to fill in. Never invent amounts, dates, bank details or tax numbers.
  6. You cannot browse, run code or see live data. Say so if asked.

Tone: warm, plain, direct.

3. Conversation starters

  • "Check my invoice before I send it."
  • "Write a friendly reminder, my invoice is 7 days late."
  • "My client has ignored two reminders. What now?"
  • "Help me set clearer payment terms for new projects."

4. Knowledge-file plan

FileContentWhy neededUpdate frequency
invoice-checklist.pdfRequired invoice fields and common omissionsConsistent checksYearly
reminder-tiers.mdThree tiers with timing and sample wordingConsistent escalationTwice a year
payment-terms-examples.mdSample late-fee and deposit clausesStarter 4Yearly

5. Refusal boundaries

  • Will not threaten legal action or write letters implying it. Says: "I can draft a firm final reminder instead."
  • Will not give tax or legal advice. Points to an accountant or solicitor for VAT, late-fee enforceability, small-claims or debt-collection questions.
  • Will not invent invoice details or edit figures without the user's confirmation.
  • Will not write messages that are harassing, deceptive or shaming.

6. Default output format

  • Diagnosis: one sentence.
  • Draft or checklist: a message in a quoted block, or a bulleted list of fixes.
  • Missing details: the placeholders to fill in.
  • Next step: one suggested action and timing.

7. Test messages

  1. "Invoice #204 for £1,800 was due
Generated from this prompt by AI. Your results will vary.

Tips and common edits

Use before opening the GPT editor. Fill in the job and the target user, then paste the instructions and starters into the builder.

Questions about this prompt

More in ChatGPT

Custom GPT Blueprint: Instructions, Knowledge Files, and Test Cases

Plan a custom GPT before building it. Produces the instruction text, knowledge file plan, conversation starters, and a test set so you can decide whether to build, narrow, or skip it.

Agent Pipeline Failure Mode Audit with Retry, Fallback, and Human Escalation Rules

Audits an existing multi-agent pipeline for failure modes and produces retry limits, fallback paths, escalation triggers, and a test plan. Supports the decision on what to harden before wider rollout.

Agent Run Pre-Mortem and Self-Correction Loop Designer

Designs verification checks, error handling, and a self-correction loop for an agent task before it runs, based on likely failure modes.

Coding Agent Task Brief with Guardrails, Test Plan, and Rollback Steps

Creates a scoped brief for an AI coding agent that limits file changes, defines tests, and includes a rollback plan and review checklist.

Trending

Developmental Edit Memo for an Essay Draft With Revise-or-Cut Decisions

Gives a draft essay a structured developmental edit: thesis check, paragraph-by-paragraph audit, and a prioritized revision plan so you can decide what to keep, rewrite, or cut before line editing.

30-Minute Moderated Usability Test Script with Tasks and Notes Template

Produces a timed 30-minute moderated usability script with task scenarios, neutral follow-up questions and a ready-to-use notes template.

4-Week Small Business Launch Plan: $0 and Small-Budget Versions

Builds a 4-week launch plan for a product or service in a specific market, with channels and weekly tasks in a $0 version and a small-budget version.

60-Minute Lesson Plan with Checks for Understanding and Differentiation

Builds a timed, objective-aligned lesson plan with formative checks, differentiation, and a teach-or-revise readiness verdict. It asks for missing inputs first and keeps assumptions separate.

90-Day Lifecycle Marketing Diagnostic and Experiment Roadmap for Onboarding and Retention

Diagnoses drop-off across the customer lifecycle from your funnel data and produces a prioritized 90-day plan of messages, triggers, and experiments with metrics so teams can decide where to invest first.

Above-the-Fold Landing Page Critique with 3 Headline, Subhead, and CTA Variants

Critiques landing page copy for clarity and conversion, then rewrites the headline, subhead, and CTA in three distinct variants.

Adaptive Quiz from My Notes with Weak-Spot Summary

Quizzes you one question at a time from your own notes, adjusts difficulty after each answer, and ends with a ranked list of weak spots.

AI Paper Explainer with Evidence Table and Practical Takeaways

Turns a pasted AI paper abstract or text into a plain-language explanation of what is new, how it was tested, its limits and what to do with it.