← Mini-Rook / Development diary

MINI-ROOK / BUILD LOG 024

The Evidence-Backed Editor

Provenance hints for release claims, with human review still required

EVIDENCE UI · SCOPED HUMAN PASS. Adapted from the project development record BUILD_LOG_024.md. Results describe the recorded checkpoint; publishing this article adds no new runtime test.

TECHNICAL CHECKPOINT

What this build actually records.

Stack and method
Release inbox lookup · metadata references · lexical overlap
Recorded progress
Syntax and isolated reporting tests pass; creator confirms source link, categories and targeted classification correction.
Still open at this checkpoint
Overlap with author-written release notes is not proof; offline, licensing, activation and links require review.

Software development on an existing workstation. This entry does not establish a new robot, custom hardware build or autonomous publishing capability.

TEST STATUS / SEPARATE FROM THE ORIGINAL RECORD

What later checks established.

Syntax and isolated reporting tests pass; creator confirms source link, categories and targeted classification correction.

Still open: Overlap with author-written release notes is not proof; offline, licensing, activation and links require review.

01 / BUILD LOG 024

The original development record

Mini-Rook now includes a release-specific Evidence Editor inside the existing Growth Draft Queue's Check claims report. It locates the selected GitHub Release from the read-only local release inbox by repository and tag. It shows the official source link and groups sentences by conservative signals: references to GitHub release metadata, lexical overlap with publisher-authored release notes, or statements that need independent human evidence.

These are provenance hints, not proof that a claim is true. The release notes are author-provided; keyword overlap can be misleading. The reviewer must still inspect offline claims, scope of free modules, activation requirements, download links and other material product statements. No public post is created or published, and the existing Quality Gate remains in place.

Verified: Python and JavaScript syntax PASS. Isolated tests confirmed report generation with metadata reference and needs-human-evidence classes, and correct non-applicability to unrelated drafts. The creator subsequently verified the real Growth UI: the correct CharacterForge GitHub source appeared, the generated post was divided into evidence categories, and an initially overconfident metadata classification for an offline/licensing claim was corrected. The mixed claim now appears as 'Release notes overlap (NOT verified)', not verified GitHub metadata. This is a human acceptance PASS for the evidence-classification UI and the targeted regression, not certification of all claims.

Status: READY TO PUBLISH — scoped UI acceptance PASS; manual factual review still required. Chapter VII — The Raven Takes Initiative.

KEEP FOLLOWING THE BIRD

The notebook stays open.