Skip to content

Instantly share code, notes, and snippets.

@peterhartree
Created January 21, 2026 15:35
Show Gist options
  • Select an option

  • Save peterhartree/219ec15687bdce13a979096a5aea79b2 to your computer and use it in GitHub Desktop.

Select an option

Save peterhartree/219ec15687bdce13a979096a5aea79b2 to your computer and use it in GitHub Desktop.
Prompt for creating comprehensive, chronological meeting summaries

Meeting summary prompt

Create comprehensive, chronological summaries that help re-envision and remember the conversation. The chronological structure is key—it allows details not explicitly in the summary to be recalled by following the flow.

Document structure

# Meeting summary: [Title]

**Date:** YYYY-MM-DD
**Participants:** [Full names with roles if relevant]

---

## Summary

[3-5 sentence overview of what the call was about and what was concluded. If listing items, use a numbered list with each item on its own line.]

---

## Part 1: [Opening topic/context]

[Chronological narrative of this phase of the conversation...]

---

## Part 2: [Next major topic]

[Continue chronologically...]

---

## Part N: The path forward / Next steps

[Clear statement of decisions and actions decided upon]

**Specific actions agreed:**
1. First action
2. Second action
3. ...

**[Person]'s next step:**
[Immediate action to take]

---

## Appendix 1: Open questions

- Question 1
- Question 2

---

## Appendix 2: Key quotes

| Speaker | Quote | Context |
|---------|-------|---------|
| Name | "Quote text" | Brief context |

---

## Appendix 3: Underlying dynamics

[Optional—include when there are important subtext, emotional dynamics, or strategic considerations worth noting. Skip if not applicable.]

**[Dynamic 1]:** Explanation...

**[Dynamic 2]:** Explanation...

Formatting guidelines

Lists:

  • Use numbered lists when items might be referred to later (options, action items, decisions)
  • Use lettered lists (a, b, c) when you don't want to imply prioritisation
  • Use bullet lists only for items that won't need to be referenced
  • Always put each list item on its own line (no inline numbered lists)

Quotations:

  • Use inline "quotation marks" for short, key phrases within prose
  • Use block quotes (>) for longer or particularly important statements
  • Preserve exact wording for:
    • Expressions of certainty/uncertainty
    • Commitments and decisions
    • Memorable or distinctive phrasing
    • Insights and realisations

Structure:

  • Use --- horizontal rules to separate major sections
  • Bold speaker names when attributing quotes or positions
  • Use tables for structured comparisons (e.g., concept assessments)

Content guidelines

Depth: Match depth to the richness of the conversation. A substantive 45-minute strategy discussion warrants 150-200+ lines; a brief check-in might need only 50.

Chronological narrative: Tell the story of how the conversation unfolded and how conclusions were reached. This helps mentally reconstruct the call.

Capture insights: Don't just list decisions—capture the reasoning, realisations, and shifts in thinking that led to them.

Appendices:

  • Open questions (Appendix 1): Almost always include; usually first appendix
  • Key quotes (Appendix 2): Include when there are memorable or important phrasings
  • Underlying dynamics (Appendix 3): Include when there's important subtext (emotional state, strategic considerations, relationship dynamics); skip when not applicable

Example part structure

## Part 2: Diagnosing the problem

Valgeir asked a pivotal question: "Is this a brief issue or an execution of the brief issue?"

This opened up a crucial realisation. Valgeir introduced the "new aesthetics" framing—the emerging visual language of rationalism, EA, and AI safety communities.

Peter's reaction was immediate recognition:
> "We should be at the front of that. In my dream world, if I was CEO, I would have put a lot into trying to be at the front of that."

**The diagnosis crystallised:** The brief hadn't communicated this ambition. Peter admitted: "It definitely wasn't pitched that ambitiously."

This explained the expectation gap: Peter was unconsciously hoping for something revolutionary while the agency was delivering competent-but-safe work as briefed.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment