threadwright

Threadwright is a framework for exploring, structuring, and accumulating knowledge while the thing being explored is still changing. It emerged from practice rather than being designed as a complete system in advance, and it addresses the problem of navigating possibility when the shape of the territory is not yet known. The framework is described as an instrument for finding out, not a machine that already knows the answer. It preserves what was tried, what was learned, what remains uncertain, which alternatives were rejected, why they were rejected, what changed, and what should be reconsidered when the context changes. The history matters because a rejected path is not necessarily a refuted path; sometimes it was simply unaffordable, the constraints changed, the evaluator changed, or the map became better. Threadwright is part of a larger expedition metaphor where humans and intelligent machines work together, with agents acting as instruments or rovers that can investigate, generate alternatives, analyze evidence, execute bounded actions, and return observations. The framework tracks where we have been, what we learned, which paths failed, what remains uncertain, and where to explore next. It is intended to chart the unknown and to support a loop of exploring, observing, learning, updating the map, reassessing, and exploring again.