Most tools measure what was said. Field measures how it landed.
Field is an open-source analysis tool that takes any text — a conversation transcript, a meeting record, a policy document, an interview, a journal entry — and analyses it across four relational dimensions.
| Symbol | Metric | Question |
|---|---|---|
| E | Energy | Who is speaking, and how much? |
| s | Symbolic Coherence | Is the language doing real work? |
| c² | Connection² | Is the signal landing clearly? |
| I | Impact | What will remain after the conversation ends? |
Impact is derived: I = (E × s) / c²
Connection is squared because poor connection degrades everything else exponentially — even brilliant, coherent content loses its effect if it doesn't land.
- Live text analysis — paste any text and get scored results instantly
- File upload — drag and drop or upload
.txt/.mdfiles - Calculation transparency — every score is fully auditable via the ℹ️ panel
- Document comparison — analyse two texts side by side with metrics, network graph, and conceptual overlap
- Speaker equity detection — use
Speaker: textformat to unlock distribution analysis
npm install
npm run devThen open http://localhost:5173.
npm run build
npm run previewField accepts plain text. For richer equity analysis, use speaker labels:
Alice: I've been thinking about how we handle feedback cycles.
Bob: Right, there's definitely a gap between what gets written up and what actually changes.
Alice: Exactly. The documentation exists but it doesn't have weight.
Without speaker labels, the text is treated as a single voice.
All analysis runs entirely in the browser — no data is sent to any server.
The scoring heuristics are intentionally transparent and adjustable. The calculation logic lives in src/utils/analyse.js.
Field is part of the Verse-ality relational intelligence framework — a framework for designing and evaluating systems that remain humane under scale.
Scores are heuristics, not verdicts.
MIT