ChatGPT/LLM Prompt: Outline Builder (Writing and Editing Books)

1. This is a single, copy/paste outlining prompt for ChatGPT (or any LLM). It takes your book idea (and optionally your Book Blueprint / brainstorm output) and turns it into a draft-ready outline you can actually write from.

It works for:

• fiction or nonfiction

• any genre (genre is optional—leave it blank for genre-agnostic)

• standalone books or series (optional series_context input)

• writers who want either a fast outline or a detailed, beat-by-beat roadmap

What makes it different

• Default output is detailed and usable: chapter list + bullet beats per chapter (not a vague summary).

• Clarifying questions are controlled by a number: clarifying_questions_count (0–10). Then it still produces an outline using reasonable assumptions (no waiting).

• Genre/trope controls are optional but supported: tropes to include/avoid + mandatory conventions.

• Strict avoid_strings support: you can ban specific words/phrases and the output will never include those exact strings.

• Planning depth control: planning_depth (0–10) lets you decide how high-level vs detailed the beats should be.

What the output looks like (plain text, copy/paste friendly)
When you run the prompt, it outputs a consistent outline in this format:

• Questions (optional; Q1..Qn)

• OUTLINE START

  • Assumptions Made (optional)

  1. Outline Snapshot (structure model, estimated chapters, promise, constraints enforced)

  2. High-Level Arc (fiction: act/phase breakdown + stakes escalation; nonfiction: reader journey + framework progression)

  3. Detailed Outline (Unit 1..Unit N)

     • For each unit:

       • Unit goal

       • Beats (bullet list)

       • Conflict/complication

       • Turn/decision

       • Ending hook

       • Continuity anchors

  4. Chapter/Section Types and Rhythm (optional but recommended)

  5. Drafting Notes (what to emphasize/avoid + what to do next)

• OUTLINE END
  If the outline is long and hits output limits, it ends with: CONTINUE