Persona / Role Prompt Generator

Build a system prompt by filling in slots: expertise area, audience, tone, output format, refusal patterns. Useful as a starting point — every output needs editing — but produces a more complete first draft than starting from a blank page.

How to use the Persona / Role Prompt Generator

Fill the slots. Output is a system-prompt skeleton you can drop into any LLM. Edit it — the value is in your specificity, not in the template. The constraints section is where you put domain-specific rules ("don't recommend hosted databases above tier X," "always include a follow-up question").

A scaffold for system prompts

A system prompt sets the model's role, audience, tone, output format, and the boundaries of what it should and shouldn't do. Starting from a blank page, it's easy to forget one of those dimensions, which is exactly when a model drifts off-tone or answers questions you wanted it to decline.

This generator fills those slots from your inputs — expertise area, audience, tone, format, refusal pattern, and extra constraints — into a structured first draft. It is a scaffold, not a finished prompt: the value lives in the specifics you add, so treat the output as a strong starting point to edit rather than something to ship verbatim.

Common use cases

  • First drafts — get a structured starting prompt instead of a blank page.
  • Consistent assistants — pin down tone and format so replies stay on-brand.
  • Scoping behavior — set a refusal pattern for requests outside the intended domain.
  • Encoding rules — capture domain constraints the model must always follow.
  • Onboarding — show teammates the dimensions a good system prompt should cover.

Frequently asked questions

Is the output ready to use as-is?
Treat it as a draft. The template covers the structure, but the real value is in the specific expertise, constraints, and examples you add afterward.

What does the refusal pattern control?
How the assistant handles out-of-scope requests — refusing firmly, acknowledging and redirecting, or no explicit pattern. It shapes behavior at the edges of the intended domain.

Where do domain-specific rules go?
In the additional constraints field — one rule per line, such as “always include a follow-up question” or “never invent function signatures”.

Does the tone setting really change output?
Yes. Tone and output-format choices noticeably steer phrasing and structure, which is why pinning them down up front keeps responses consistent.
Embed this tool on your site

Free to embed, no attribution required (but appreciated). Paste this where you want the tool to appear: