fixaiprompt
All templates
⚡ Advanced Techniques·Works on: claude, chatgpt

Skeleton-of-Thought Outline-then-Fill

Technique: Skeleton-of-Thought

Generate the outline first. Then expand each section in parallel. Fast and consistent on long generations.

Advanced#skeleton#long-form#writing
0/1
Fill the template
1 placeholder left.
Live preview
Use Skeleton-of-Thought. Don't write linearly — build the skeleton first.

Deliverable:


**Stage 1 — Skeleton.**
Produce a numbered outline. Each outline item is ONE LINE with:
- Section title.
- The single point this section makes.
- The evidence or example it uses.

Do NOT write any section content yet.

**Stage 2 — Skeleton review.**
Look at the outline:
- Does it flow? Each section should set up the next.
- Is anything missing? Is anything redundant?
- Is the order optimal for reader retention (most-interesting payoffs not at the end)?

Fix the outline if needed before continuing.

**Stage 3 — Fill in parallel.**
Now write each section IN ISOLATION. Do not reference what came before or after — each section should be a self-contained mini-document that hits its single point. Length proportional to importance.

**Stage 4 — Smooth.**
Write ONE transition sentence between adjacent sections so the seams don't show.
Inputs0 of 1
›See the lazy version this template replaces
Before — the lazy prompt
Write <long deliverable>.

Why it works

  • Skeleton-first prevents the model from over-investing in the first section and running out of momentum.
  • Parallel filling is consistent because each section answers a single question — no drift.
  • Outline-then-review catches structural issues cheaply (before any prose is written).
  • Transition sentences as a final pass produce flow without padding.

Make this one yours

Replace the bracketed placeholders, then paste into the Prompt Fixer to lint your customisation before hitting send.

Open Prompt FixerBrowse more templates