Skeleton-of-Thought
Reasoning
Level 2 · Advanced reasoningGenerate a skeleton (outline) of the answer first, then expand each point — better organization, especially for long-form output.
Skeleton-of-thought asks the model to first produce a short outline of the key points the answer will cover, then expand each point into full content. Separating 'what are the parts' from 'what does each part say' tends to produce better-organized long-form answers than a single continuous pass.
When to use it
Use it for long-form answers with multiple distinct parts — reports, explanations, comparisons — where a single continuous pass risks skipping or under-developing some points.
When to avoid it
Avoid it for short, single-idea answers where there's nothing meaningful to outline.
Advantages
- Produces more consistently organized long-form answers
- Makes it easy to spot a missing or underdeveloped point before the full answer is written
- Each point tends to be expanded to a comparable, balanced length
Disadvantages
- Adds an extra step compared to a direct answer
- If the initial skeleton is itself incomplete or poorly scoped, the expansion inherits that gap
Examples
Write an explanation of how vaccines work, for a general audience. First, produce a short outline of the 4-6 key points this explanation should cover. Then expand each point into a few clear sentences, in order.
Pairs well with
Try it
Fill in the placeholders below and see the compiled prompt update live.
What you want to achieve.
- 25
- Words
- 149
- Characters
- 0/1
- Placeholders filled
- 1 min
- Reading time
- Completeness0
- Clarity0
- Context0
- Specificity0
- Constraints0
- Format0
- Output0
- Constraints note0
- Success criteria0
- Scope focus100
- Verification step0
- Negative example0
- Confidence0
- Placeholder syntax100
- Markdown structure0
- Fill in the remaining placeholders.
- Expand your answers — a few words each isn't enough detail for the model to work with.
- Add background context so the model understands the situation.