Research records

A technology test should not end as an issue comment.

Dated, attributable records of Masters research results, technology comparisons, prior-art reviews, negative results and architecture decisions. Each record keeps its method, its limitations, its negative results, and the conditions that would cause Masters to reconsider it.

Git-managed source of truth Dated and attributable Negative results published

What counts as a Finding.

A Finding is a publication object: an evidence-backed result from a test, prior-art review, comparison, implementation experiment, negative result, or architecture decision. An essay is authored interpretation and is labelled as one. A method record describes how Masters works. A specimen illustrates the shape of that work and is not a client case study. The four share this pipeline and never share a truth boundary.

Where a record depends on an evidence-bearing Masters claim, it names that claim by its stable identifier instead of restating it, and the published page is rendered from the claim record itself. Correcting the claim and rebuilding corrects every record that cites it.

Finding · Published

A Git-native browser editor for Masters Findings: one candidate wins the workflow, the other wins the credential

Both candidate editors represent the Masters publication contract faithfully and lose nothing in a round trip. Only one produces a pull request automatically — but only the other can hold a credential scoped to a single repository, and the branch protection the first one depends on is not available on this repository at all.

MCG-F-0002 · Published 2026-09-05 · Confidence: provisional