RFR-2026-016: Complete blinded mechanism-boundary adjudication

Research opportunity

Complete or formally supersede the controlling EX-FE-0002 Stage A and Stage B mechanism-boundary experiment.

Background

The 2026-07-28 source synthesis narrows the working model, but EX-FE-0002 remains Stage A incomplete, Stage B blocked, and experimentally inconclusive.

Origin documents and trace

• research/framework-engineering/experiments/EX-FE-0002/START-HERE.md, section: Two-stage execution; Stage A acceptance gate; Decision rules
• research/evaluations/FE-BOUNDARY-2026-07-24-v1.1/stage-a-decision.md, section: Stage A decision — v1.1
• research/evaluations/FE-BOUNDARY-2026-07-28/RP-FE-BOUNDARY-2026-07-28.md, section: F8; Compatibility with existing experiment authority

Specific assumption challenged: A same-provider source synthesis can support a working classification without completing the controlling independent experiment.

Supporting evidence: EX-FE-0002 defines the governing blinded test; its v1.1 decision explicitly blocks Stage B, and the successor REP says it does not override that state.

Unknowns

• Whether any FE mechanism survives blinded lossless-subsumption review
• Whether source cards pass independent completeness audit
• Whether blinding and recognition thresholds can be met

Dependencies

• RFR-2026-001

Suggested REP and methodology

Create REP-2026-016. Use the new source registry as candidate repair input; appoint disjoint curators, completeness auditor, blinding editor, recognition coordinator, key custodian, reviewers, adjudicator, and analyst; freeze and hash artifacts; execute Stage B only after every gate passes.

Expected outputs

• Accepted or rejected Stage A source cards
• Blinded reviewer packet and sealed-key receipt
• Adjudicated mechanism-boundary decision

Success criteria

Every protocol gate is evidenced, the result remains null/negative/inconclusive when warranted, and the accepted decision either identifies a non-subsumed operational mechanism or narrows/supersedes the claim.

Execution recommendation

• Recommended agent: independent mechanism-boundary research director
• Estimated effort: high
• Expected knowledge gained: Resolves the controlling experimental state and defines which components, if any, merit causal utility testing.
• Frontier Score: 492
• Score inputs: knowledge gain 5/5; impact 5/5; cross-project reuse 5/5; scientific importance 4/5; dependency cost 4/5; implementation difficulty 4/5.
• Score calculation: 5 × 5 × 5 × 4 − 4 − 4 = 492.
• Score status: structured expert judgment, not measured utility.