Research frontier handoff

Objective

Maintain an auditable, traceable, dependency-aware inventory of where the repository's accepted/current knowledge ends and which research should move the boundary next.

Current result

• 1246 publishable records have explicit dispositions.
• 40 accepted/current or authority-dependent artifacts are analyzed through 11 evidence-coherent units and individual document frontiers.
• 16 open RFR candidates remain after semantic merging.
• RFR-2026-001 is the first funded REP recommendation.
• RFR-2026-008 can begin in parallel.
• RFR-2026-016 resolves the controlling EX-FE-0002 experimental state before RFR-2026-003 tests causal utility.

Authority

RFRs are open candidates and mutable. This analysis does not accept them, promote theory, supersede EX-FE-0002, or authorize autonomous canonical mutation. Acceptance requires a named governance decision; accepted changes then use successor and supersession records.

Reproduction

npm run research:build
node research/frontier/build-frontier.mjs
node research/frontier/build-frontier.mjs --check
npm run research:validate

The builder checks source existence, accepted/current coverage, duplicate IDs, score reproduction, rank order, dependency integrity, cycle freedom, unit top-five completeness, and generated-output staleness.

Revalidation triggers

• ECR-000003 review closes.
• EX-FE-0002 completes or is superseded.
• The boundary decision is accepted, rejected, or superseded.
• ROS–FE v1.1 lands.
• An RFR changes status.
• A top-four RFR completes.
• The publisher catalog changes accepted/current membership.

Next action

Create the REP for RFR-2026-001. In parallel, prepare the v1.1 proposal and negative fixtures for RFR-2026-008. Do not start RFR-2026-003 until both its human-baseline and controlling mechanism-boundary dependencies are satisfied or explicitly superseded.