Extract independent ideas from a source article and adapt them to different short-form contexts
8 min setupTested with:ChatGPTReviewed: 2026-08-28
1
Add your context
Your text stays in this browser. AILesson Prompts does not send it to a model or server.
2
Your prompt
Unfilled fields remain visible as placeholders, so you can still copy and edit the prompt
Repurpose the supplied article into genuinely distinct short-form pieces.
Complete article, headings, links, claims, examples, and rights:
[article]
Platforms, count, format, length, audience, objective, CTA, and schedule:
[outputs]
Voice, invariant facts, quotation, links, disclosure, exclusions, and review:
[rules]
Create a source map of claims, examples, caveats, and section locators. Select ideas that can stand alone without losing essential context. Give each piece one primary audience question and a different angle; do not merely shorten the introduction repeatedly. Adapt structure, pacing, opening, terminology, CTA, and accessibility to its platform while preserving meaning, uncertainty, attribution, and causal boundaries. Do not invent transitions that imply new facts, personal experience, quotes, outcomes, or consensus. When compression would make a claim misleading, omit it or retain the necessary qualifier. For every draft provide source mapping, what was intentionally omitted, and any required link or disclosure. Audit duplication across drafts, factual fidelity, character limits, CTA fit, and whether each piece remains understandable without the article.
Private by defaultPrompt assembly happens locally in your browser. Avoid placing confidential information into any AI service unless your organization allows it.
From input to outcome
A worked example
See how concrete context turns this recipe into a usable result
Actual input
Source article
Title: Why status visibility does not equal ownership. Section 1 defines visibility as seeing state and ownership as authority to decide and accept. Section 2 gives a fictional handoff example labeled fictional. Section 3 reports 7 of 10 anonymized interviewees mentioned unclear ownership; convenience sample, not prevalence. Section 4 offers a four-question audit. Conclusion: tools can support but not assign accountability. We own republishing rights.
Target outputs
Three pieces: LinkedIn post under 1,000 characters for operations leads, email tip under 180 words for existing subscribers, and 6-frame carousel outline for team leads. Each should lead to the full article, published across one week.
Adaptation rules
Direct, calm voice. Keep fictional label and 7-of-10 sample caveat. No direct quotations. Use placeholder <article link>. No claim that tools cause failure or that the issue is universal.
Example output
LinkedIn — distinction angle
A green status can tell you where work is. It cannot tell you who has authority to accept it. Visibility and ownership solve different questions: “What state is this in?” versus “Who decides it can move?” In a convenience sample of 10 interviewees, 7 mentioned unclear ownership; that small sample is a signal to investigate, not a prevalence estimate. Try the article's four-question handoff audit: <article link>
Source: sections 1, 3, and 4. Omitted the fictional example to keep one idea.
Email tip — practical audit angle
Before adding another status field, inspect one recurring handoff. Ask who declares work ready, who accepts it, what evidence counts as accepted, and who resolves disagreement. A tool may display these decisions, but it cannot responsibly make them for the team. Read the full explanation and audit: <article link>
Source: section 4 and conclusion. Omitted the interview number because it is not needed for the action.
Carousel — scenario angle
1 Visibility is not ownership. 2 You can see “ready” and still lack an accepter. 3 Fictional example: a report waits because sender and reviewer use different completion rules. 4 Name the deciding role. 5 Define acceptance evidence and escalation. 6 Audit one handoff, then choose tool support: <article link>.
Source: sections 1, 2, 4, conclusion. Frame 3 retains the fictional label. No draft claims universality or that tools cause failure.
Why this works
1
Standalone idea selection prevents context-dependent claims from becoming misleading snippets.
2
Cross-draft duplication checks create a portfolio rather than repeated summaries.
Check the result
Can each draft stand alone with essential context and qualifiers?
Does every claim map to a source section without changed meaning?
Are angle, structure, and CTA genuinely adapted by platform?
ChatGPT