2026 · 3D PRINTING

PartMatch

Killed early, with the avoided build cost treated as the result.

SCRAPPED Ran the adversarial PRD review before a build began.
On this page

CONTEXT

The work started here.

The idea proposed turning photographs of broken objects into printable replacement-part geometry.

BUILD

What was built.

3D-printable replacement parts from photos of broken ones. It exists as a PRD and a no-go memo.

PROJECT DESIGN

How it was shaped.

The adversarial PRD review asked one question the plan couldn't answer: had the hard part, turning a photo into printable geometry, actually been proven feasible? It hadn't.

LESSONS

What changed afterward.

  • “Spike first” is the cheapest sentence in product development: stopping it cost one review session instead of a build month.

WHY IT STOPPED

The hard technical premise was unproven.

The review could not establish that photo-to-printable geometry was feasible, so the project stopped at a PRD and no-go memo.

CONTROL

The constraint that held.

A feasibility spike is required before committing a build month to an unproven technical core.

EVIDENCE

Where to inspect it next.

Build registerReturn to the build register

CARRIED FORWARD

What survived the project.

Adversarial review as a standing gate before any build.

WRITING

Field notes from this work.

No public field notes for this project yet.

Trust belongs in the schema, not the application.
The operating principle behind every project here. A rule the database enforces can't be forgotten in a hurry. Constraints, denied-by-default access, append-only logs — the controls that hold are the ones the system won't run without.