# 📌 Prompt: Zoom Chat → Condensed Signal Map (Questions, Observations, Social Dynamics, Resources)

You are given a **Zoom chat transcript** (plain text or exported chat), containing a mixture of:

* Questions (explicit and implicit)
* Observations or assertions
* Reactions and social signals (e.g. “+1”, agreement, humor, frustration)
* Meta-process comments (about the exercise, tools, or facilitation)
* Links and references (articles, tools, documents, videos)
* Short fragments or half-formed thoughts
* Repeated or slightly varied contributions from multiple participants

Your task is to **extract and lightly condense the *substantive signals*** from the chat, **without answering, summarizing themes, or editorializing**.

Focus on preserving **what participants noticed, questioned, felt, proposed, and shared**.

---

## Step 1 — Extract (verbatim-leaning)

* Parse the chat **line-by-line**.
* Strip timestamps, usernames, and system messages **unless they affect meaning**.
* Preserve original wording as much as possible.
* Treat fragments, informal phrasing, and unfinished thoughts as valid inputs.
* Extract **links and referenced resources exactly as posted**.

---

## Step 2 — Classify (do not answer or interpret)

Classify each extracted line into **one primary category**:

### Core Content

* **Question**
  (Explicit or implicit; “what does this mean?”, “are we…?”, “how might…?”)

* **Observation / Claim**
  (Assertions, interpretations, pattern-noticing, “this feels like…”, “I notice…”)

* **Concern / Uncertainty**
  (Confusion, worry, overwhelm, doubt, unease)

* **Proposal / Suggestion**
  (Ideas, requests, “what if we…”, facilitation or process changes)

* **Constraint / Limitation**
  (Technical, cognitive, accessibility, time, attention, capacity)

### Meta & Social

* **Process / Meta-Process Signal**
  (Comments about the exercise, tools, pace, facilitation, or inquiry itself)

* **Social / Emotional Signal**
  (Empathy, frustration, humor, appreciation, alienation, reassurance)

### References

* **Resource / Reference**
  (Links, books, articles, tools, people, concepts; include URL if present)

* **Historical / Analogical Reference**
  (Invoked past events, figures, eras, metaphors)

* **Noise**
  (Emoji-only, “+1”, “thanks”, link reposts with no commentary)

Do **not** merge categories at this stage.

---

## Step 3 — Light Condensation (signal-preserving)

* **Fold clearly duplicated or near-duplicate items** into a single broader item.
* When multiple people express the *same idea* in different words, merge cautiously.
* Preserve **distinct phrasing** if it adds nuance or angle.
* Attach repeated reactions implicitly to the merged item.
* Normalize wording **slightly** for clarity, but:

  * ❌ Do not abstract into themes
  * ❌ Do not smooth away emotional or awkward phrasing
  * ❌ Do not resolve ambiguity

---

## Step 4 — Preserve Outliers & Social Texture

* Keep **odd, sideways, speculative, or emotionally charged** inputs.
* Do not discard content because it was not answered, discussed, or resolved.
* Preserve signals about **energy, mood, overwhelm, trust, or disengagement**.
* Keep humor, irony, or absurdity **if it played a social or grounding role**.

---

## Step 5 — Structured Output

Produce a structured result with **clearly labeled sections**:

1. **Condensed Question Set**
   (Questions only, deduplicated, clearly phrased)

2. **Key Observations & Claims**
   (Non-questions that assert or notice something about the situation)

3. **Concerns, Constraints & Uncertainties**
   (Barriers, limits, worries, confusion)

4. **Proposals & Process Suggestions**
   (Ideas, requests, facilitation changes)

5. **Social & Emotional Signals**
   (Mood, appreciation, frustration, humor, disconnection)

6. **Resources & References Shared**
   (Links, books, tools, articles — include URLs)

7. **Notable Outliers / Edge Inputs**
   (Weird, provocative, personal, or orthogonal contributions)

8. **Discarded Noise (listed, not hidden)**
   (Emoji, thanks, +1s, reposted links without commentary)

---

## Optional: Layered Interpretation (only if it emerges naturally)

If the content supports it, produce a **layered view**:

* **Surface** — immediate questions, reactions, logistics
* **Structural** — systems, incentives, constraints, coordination dynamics
* **Civilizational / Strategic** — long-term norms, power shifts, meaning systems

Do **not** invent layers. If unclear, omit this section.

---

## Hard Constraints (important)

* ❌ Do **not** answer any questions
* ❌ Do **not** summarize abstract “themes”
* ❌ Do **not** judge correctness or quality
* ❌ Do **not** privilege expert voices over others
* ✅ Prefer completeness over elegance
* ✅ Preserve the **voice, uncertainty, and social texture** of participants
* ✅ Treat the chat as a **sense-making artifact**, not a debate transcript