FESADev/phases/1-linear-static-mitc4/step15.md

Step 15: phase1-evaluator-closeout

Sprint Contract

Objective

Run the full Phase 1 evaluator pass, close documentation and handoff gaps, and record Phase 2 carryovers.

Required Reading

• /AGENTS.md
• /PROGRESS.md
• /PLAN.md
• /docs/README.md
• /docs/HARNESS_ENGINEERING.md
• /docs/PRD.md
• /docs/ARCHITECTURE.md
• /docs/ADR.md
• /docs/NUMERICAL_CONVENTIONS.md
• /docs/ABAQUS_INPUT_SUBSET.md
• /docs/VERIFICATION_PLAN.md
• /docs/RESULTS_SCHEMA.md
• /docs/MITC4_FORMULATION.md
• /references/README.md
• /phases/1-linear-static-mitc4/index.json

Scope

• Evaluate all Phase 1 Definition of Done items from PLAN.md and docs/PRD.md.
• Run the full validation command set and inspect reference-comparison evidence.
• Produce a concise evaluator report in PROGRESS.md or a linked docs artifact.
• Move unresolved Phase 2 or deferred work into PLAN.md.

Allowed Files

• docs/**
• tests/** only for evaluator-owned missing validation metadata, not feature implementation
• references/** only for documentation or final accepted artifact notes
• PLAN.md
• PROGRESS.md
• phases/1-linear-static-mitc4/index.json

Explicit Non-Goals

• Do not implement missing solver features in the evaluator step.
• Do not mark Phase 1 complete if validation, reference comparison, MITC4 formulation, RF recovery, or parser rejection evidence is missing.
• Do not hide remaining risks in prose without putting future tasks in PLAN.md.

Tests To Write First

• No new implementation tests are expected unless the evaluator identifies missing metadata tests that are needed to prove acceptance.

Reference Artifacts

• Required: all accepted Phase 1-compatible reference cases and their *_displacements.csv files.
• Optional: *_reactions.csv; otherwise evaluator must verify RF through internal equilibrium tests.

Acceptance Commands

python scripts/validate_workspace.py

Evaluator Checklist

• Phase 1 Definition of Done is checked item by item.
• Parser rejects unsupported Abaqus features with tests.
• MITC4 baseline tests and documented formulation agree.
• U reference comparison evidence exists for accepted cases.
• RF is recovered from full vectors and verified by available reference or equilibrium tests.
• PLAN.md and PROGRESS.md are current.

Handoff Requirements

• Record final verdict, validation output, reference evidence, and remaining risks in PROGRESS.md.
• Move future work, Phase 2 candidates, and deferred features into PLAN.md.
• Update the matching phase index entry: use completed with a one-line summary, or blocked with blocked_reason, or error with error_message; update the top-level phase status if execution is complete.

Verification

1. Run the acceptance command.
2. Perform the evaluator checklist against changed files and test evidence.
3. Confirm no phase step remains with stale status or missing handoff notes.

Do Not

• Do not use evaluator closeout as a place to sneak in broad implementation fixes.