FINAL TRANSVERSAL ARCHITECTURAL AND
                        METHODOLOGICAL ALIGNMENT
MENS-SCCA × APEX — Final integration note

1. SCOPE AND EPISTEMIC STATUS
This note closes the architectural alignment of the proposal. It does not introduce a new research line, replace the established
protocol, or alter the two parallel research paths. It specifies the common transversal structure through which the existing
methodology is interpreted. MENS-SCCA remains research infrastructure and methodological instrumentation; APEX remains the
external professional-agent evaluation environment and benchmark.

2. TWO PARALLEL PATHS AND ONE TRANSVERSAL CONDITION
 PARALLEL PATH A — EXTENSION             Research directly aligned with Mercor’s line of advancing knowledge about evaluation of professional
 OF APEX                                 agents. The contribution is an experimental layer for investigating whether structured longitudinal
                                         representation yields incremental, reproducible information beyond existing outcome- and
                                         trajectory-based evaluation signals.

 PARALLEL PATH B — KNOWLEDGE             Research uses the same infrastructure to generate generalizable knowledge about transfer of
 FOR INTERSECTORAL ADAPTATION            procedures, evidence, and execution characteristics across professional domains. The scientific
                                         product is the benchmark/protocol and empirical knowledge generated by the research, not
                                         MENS-SCCA as a Mercor software product.

 TRANSVERSAL LINE —                      Every relevant experimental claim must remain linkable, subject to authorization and confidentiality, to
 TRACEABILITY AND                        task, version, trajectory, evidence, rubric, evaluation, and provenance.
 REPRODUCIBILITY



3. COMMON TRANSVERSAL EXECUTION CHAIN
Task → Environment → Agent → Actions/Tools → State Changes → Evidence → Trajectory → Recovery → Final Output
→ Evaluation

This is an analytical mapping, not a claim that Mercor lacks any of these objects. Mercor’s current public methodology explicitly
evaluates final work and trajectories and describes analysis of intermediate actions, tool use, external changes, error recovery,
and constraints. The research question is therefore incremental: whether an additional structured representation of these
observable elements can expose reproducible information not fully captured by existing evaluation signals.

4. ROLE SEPARATION
 APEX / Mercor                 Professional benchmark, realistic tasks/environments, authorized executions, existing outcome/trajectory
                               evaluation, expert criteria, and benchmark governance.

 MENS-SCCA                     Research instrumentation and analytical representation for longitudinal states, transformations, provenance,
                               evidence, failure propagation, recovery, consistency, and transfer.

 Research                      Empirical test of whether the added representation provides incremental, stable, reproducible, and interpretable
                               information.


5. CENTRAL RESEARCH FORMULATION
Given that APEX-Agents already evaluates final deliverables and agent trajectories, the proposed research investigates whether
an additional structured representation of trajectory, state, provenance, transformation, recovery, and longitudinal dependencies
can yield incremental and reproducible information beyond the existing evaluation signals.

6. NON-CHANGES / FINAL BOUNDARIES
• No replacement or superiority claim is made regarding APEX, Mercor, or existing evaluation infrastructure.

• No claim is made that trajectory evaluation is absent from APEX.

• No new hypotheses, metrics, or research program is introduced by this note.

• Prior MENS-SCCA architecture is treated as methodological antecedent, not as external validation.

• Confidentiality, authorization, provenance, contamination control, and separation of research infrastructure from Mercor
materials remain governed by the established data plan.

• The two parallel paths remain distinct while sharing the same traceability and reproducibility condition.
7. CLOSURE
This alignment is the final transversal closure of the proposal. It preserves the established methodology while making explicit the
coherent common structure connecting the APEX extension path, the intersectoral knowledge-generation path, and the
traceability/reproducibility condition.