Separate what is claimed, what is assumed and what people value. Learn with the example,
then prepare your own question. Free, no account; nothing is autosaved.
Five-minute exercise: how to read this example
Compare N1's reported global estimate with N2's research assessment. Neither supplies
this corridor's local data.
Open N4 and N5. Which assumptions would you test before trusting the predictions?
Open N6 and N7. What priorities and affected groups need discussion?
Open N8. Decide what you would ask next before agreeing to any proposal.
These prompts teach distinctions. There is no mastery score, automated verdict or claim
that the proposal is correct.
Guided case studyHypothetical cycle-lane pilot · Sources reviewed 8 Sep 2026
Should our city run a 12-month pilot replacing curbside parking on one side of a busy
commercial street with a protected cycle lane?
✦
No verdict. This is one hypothetical corridor, not a real
consultation. No local data has been supplied.
Claim map · 8 nodes Editorial teaching example
Discussion brief · Draft 01
The evidence narrows the argument. It does not make the choice for us.
This sample is organized by what is known, what remains local, and what must be
decided as a matter of values.
✓
What the sources inform
Global context informs the question, not this corridor’s answer.
WHO estimates that vulnerable road users account for more than half of global road
deaths.
The IPCC assesses that walking and cycling infrastructure can support
lower-emission travel.
?
What remains unknown here
Local conditions determine whether the expected benefit is plausible.
Trip length and network continuity
Current curb users and time patterns
Baseline safety and access measures
⚖
What values decide
How should safety and mobility be weighed against general parking?
Evidence can make the trade-off clearer. It cannot remove the need to choose.
↗
What a pilot can test
A reversible design can turn assumptions into public evidence.
Publish baseline measures, protect essential access and pre-agree evaluation
criteria. Before/after changes alone do not establish causation; comparison
locations and seasonal patterns matter.
Source trail
The exported brief contains every claim, its type, sources, limitations and
connections. Source confidence is not this pilot’s probability of success.
Your material · not independently checked
Prepare your own discussion.
Start with a question and two or three claims. Add source records where relevant, make
assumptions and values visible, then keep the next questions with your brief. Claim types
are your labels—not proof.
Up to 8 claims, 12 source records and 16 relationships. Changes remain drafts until
Apply. Removing a referenced item requires unlinking it first; the app never silently
removes its connections.