Motivation

Why this matters

Public argument is badly structured. Claims are vague, debate doesn't accumulate, evidence doesn't map cleanly to the specific sub-point it bears on. Discussion scatters across forum posts and never compounds into shared understanding.

The vision: take a top-level claim, recursively decompose it into sub-claims, and build a directed graph (DAG) of claims + evidence. Documents come and go, but the claim tree is a stable scaffold that discourse accumulates against — "cathedral building" instead of scattered noise.

Rufus's thread: issue trees / SCQH / Minto pyramids — wanted this tool for 10+ years. David's thread: the "Abductio" proposition-decomposition process inside his Promise Protocol.

Problem framing (SCQH per active stream)

We use SCQH — Situation, Complication, Question, Hypothesis — to sharpen each stream into a crisp problem statement. Stream definitions and the plan of work live in PLAN.md.

Stream 1 — Product / prototype

S — Static demo makes the vision resonate; the process is well specified.
C — Demo is scripted; AI does none of the work, so nothing is proven about feasibility or output quality.
Q — Can AI reliably (a) generate a useful claim tree from input, and (b) map document fragments to the right nodes — well enough to convince us — on the toy example?
H — Build two thin AI skills against the gold toy case; if output quality is good, wrap a minimal interactive UI. Toy first, real applications later.

Stream 2 — Narrative / comms

S — David + Rufus share the vision; collaboration channel + repo exist.
C — The "why this matters / why now / why us" isn't written down; David has only fragments. The four-step process has no clean shareable diagram.
Q — What's the crisp shared story (background + four-step process) that aligns David and attracts collaborators?
H — Draft a background post (Rufus's 10-yr thread: SCQH / Minto / issue trees + personal story) and a clean Mermaid four-step diagram.

Deferred / dropped

Stream 3 — Second Renaissance LTP — deferred; it's an application, too complex until a toy example works end-to-end.
Stream 4 — Discord bot wiring — dropped this round.