Claude-optimized prompt structure shaped for chatgpt.
# Task
Distill the source material into an executive brief. Every sentence must pass the "so what?" test — if it doesn't inform a decision, cut it.
# Source material{source_material}# Context- Audience: {audience}- Goal: {goal}- Decision context: {decision_context}- Constraints: {constraints}# Anti-hallucination rules1. Use ONLY information from the source material.
2. Label assumptions explicitly as [ASSUMPTION].
3. Do not recommend actions unsupported by the source evidence.
4. If evidence is weak or missing for a key point, say so.
# Brief structure1. **Bottom line up front** (1-2 sentences): The key finding or recommendation. Lead with the answer, not the context.
2. **Context** (2-3 sentences): Only enough background for the audience to understand the recommendation. No history lessons.
3. **Key findings** (3-5 bullets): Most important facts, each with supporting evidence.
4. **Risks and concerns** (2-3 bullets): What could go wrong or what's uncertain.
5. **Recommendation** (1-2 sentences): Clear, specific, actionable.
6. **The ask** (1 sentence): What you need from the audience (a decision, resources, approval, alignment).
7. **Next steps** (2-3 bullets): Who does what by when.
# Formatting rules- Total brief: under 400 words. Executives won't read more.
- No jargon unless the audience expects it.
- Bold the single most important number or finding in the brief.
# Output
Return the complete brief using the structure above. Include a word count at the end.
claudegrounded_factsthinking_medium
claude variant
Claude-optimized prompt structure shaped for claude.
<source>{source_material}</source><context><audience>{audience}</audience><goal>{goal}</goal><decision_context>{decision_context}</decision_context><constraints>{constraints}</constraints></context><task>Distill the source into an executive brief. Every sentence must pass the "so what?" test.</task><instructions>
Use ONLY source material. Label assumptions [ASSUMPTION]. Do not recommend unsupported actions. Flag weak evidence.
Structure:
1. Bottom line up front (1-2 sentences)
2. Context (2-3 sentences, minimal)
3. Key findings (3-5 bullets with evidence)
4. Risks (2-3 bullets)
5. Recommendation (1-2 sentences)
6. The ask (1 sentence)
7. Next steps (2-3 bullets: who, what, when)
Under 400 words. Bold the most important finding.
</instructions><output>
Return the complete brief with word count.
</output>
geminigrounded_factsthinking_medium
gemini variant
Claude-optimized prompt structure shaped for gemini.
Source material:
{source_material}
Audience: {audience}
Goal: {goal}
Decision context: {decision_context}Constraints:{constraints}
Rules: only source material, label assumptions [ASSUMPTION], flag weak evidence. Every sentence must pass "so what?" test.
Structure: Bottom line (1-2 sentences) > Context (2-3 sentences) > Key findings (3-5 bullets) > Risks (2-3 bullets) > Recommendation > The ask (1 sentence) > Next steps. Under 400 words.
Based on the entire content above, return the complete executive brief with word count.
Related templates
More templates in the same working category.
Meeting Summary
Summarize meetings with structured decisions, action items, disagreement tracking, and confidence ratings.