Business Subject Eval-Ready Design

Goal

Prepare business as the first post-accounting subject expansion by moving it from a deferred candidate shell to eval_ready only after a business-owned fixture pack, manifest-backed signals, and near-miss guards are in place.

This is not a runtime activation slice. Business must remain unavailable for default runtime subject switching until a later runtime-enabled promotion review proves precision after eval-ready.

Current Baseline

On dev as of July 5, 2026:

• Accounting, finance, and economics are runtime_enabled.
• The subject expansion onboarding contract is merged.
• Business is a deferred candidate with:
  • empty signal_groups,
  • empty method_lenses,
  • blank evaluation_pack.
• Business resources already exist:
  • content/skills/domain-profiles/business-management.yaml,
  • content/subjects/business/skills/business-journal-positioning-auditor.md,
  • venue profiles for Academy of Management Journal, Organization Science, Journal of Management, Journal of Marketing, and Strategic Management Journal,
  • overlays for manuscript architecture, study design, and statistics.

Current business eval-ready gate should fail closed with explicit onboarding blockers:

business --gate eval-ready:
  activation_status: candidate
  eligible_for_eval_ready: false
  blocking_failures:
    - activation_status is candidate
    - missing evaluation_pack for deferred subject
    - missing signal dimension: method
    - missing signal dimension: data_or_outcome
    - missing signal dimension: venue
    - missing signal dimension: theory_or_construct
    - missing clear_positive fixtures
    - missing method_only_borrow fixtures
    - missing near_miss fixtures

Decision

Add a business eval-ready pack as a guarded subject expansion slice.

Why business is the next candidate:

• Business resources and venue profiles already exist.
• Business has a well-defined quality bar in the local domain profile: management theory contribution, construct clarity, method transparency, and venue positioning.
• It is adjacent enough to finance/economics/accounting that near-miss guards are useful, but the intended scope can stay bounded to scholarly business, management, organization, strategy, marketing, and operations research.

Scope

In scope:

• Add a business-owned fixture pack under tests/fixtures/subject_router_eval/business/.
• Add business manifest signal groups for method, data/outcome, venue, and theory/construct dimensions.
• Add business method lenses that borrow business review guidance without switching the primary subject from one method phrase alone.
• Move business from candidate to eval_ready.
• Point business evaluation_pack at tests/fixtures/subject_router_eval/business.
• Add contract and router-gate tests proving business passes eval-ready and fails runtime-enabled.
• Re-run accounting, finance, and economics runtime gates to guard against adjacent-subject regressions.
• Update the roadmap to mark business eval-ready as the current Stage 4 slice.

Out of scope:

• Do not promote business to runtime_enabled.
• Do not add political economy, geoeconomics, or economics-accounting fixtures.
• Do not broaden accounting, finance, or economics signals.
• Do not change provider configuration, literature search, full-text retrieval, Zotero behavior, local-agent execution, or release automation.
• Do not add network-dependent verification.

Business Subject Boundary

Business can become the primary subject in eval-ready measurement only when a request has at least two business-specific dimensions such as:

• business method plus management theory construct,
• venue plus construct or theory contribution,
• organizational/firm/team/platform setting plus mechanism or construct,
• marketing/strategy/organization venue plus empirical design.

Business must not activate from:

• generic "business plan", "small business", or startup operations wording,
• course case assignments or teaching cases without research contribution,
• consulting, market analysis, sales enablement, or product marketing copy,
• project management or workflow operations,
• generic firm, customer, or market mentions that are actually finance, economics, accounting, or core planning tasks.

Method-only business evidence may still borrow business lenses without suggesting business.

Proposed Manifest Signals

The implementation plan should encode these signal groups in content/subjects/business/runtime-subject.yaml.

Method signals

Suggested method-only lenses:

• business.method.gioia
  • value: gioia-method
  • activation: method_only
  • patterns: Gioia, first-order concepts, second-order themes, aggregate dimensions
  • near misses: generic "concept map" or non-research theme sorting
• business.method.case-study
  • value: case-study
  • activation: method_only
  • patterns: multiple case study, Eisenhardt, Yin case study, within-case, cross-case
  • near misses: teaching case, business case, case interview
• business.method.process-research
  • value: process-research
  • activation: method_only
  • patterns: process research, temporal bracketing, event timeline, turning points
  • near misses: process improvement workflow

Data or outcome signals

Suggested subject-level data/outcome signals:

• business.data.organization-panel
  • value: organization-panel
  • activation: subject
  • patterns: firm-level panel, organization-level data, team-level data, manager survey, employee survey
• business.data.qualitative-fieldwork
  • value: qualitative-fieldwork
  • activation: subject
  • patterns: interviews with managers, fieldnotes, organizational ethnography, case evidence database, archival documents
• business.data.market-platform
  • value: market-platform
  • activation: subject
  • patterns: platform marketplace, customer journey, marketing channel, firm-customer interaction

Venue signals

Suggested context-only venue signals:

• business.venue.amj
  • value: academy-of-management-journal
  • patterns: Academy of Management Journal, AMJ
• business.venue.organization-science
  • value: organization-science
  • patterns: Organization Science
• business.venue.journal-of-management
  • value: journal-of-management
  • patterns: Journal of Management
• business.venue.journal-of-marketing
  • value: journal-of-marketing
  • patterns: Journal of Marketing
• business.venue.strategic-management-journal
  • value: strategic-management-journal
  • patterns: Strategic Management Journal, SMJ

Theory or construct signals

Suggested subject-level theory/construct signals:

• business.construct.theory-contribution
  • value: theory-contribution
  • activation: subject
  • patterns: management theory, theory contribution, literature stream, construct clarity, boundary conditions
• business.construct.organization-mechanism
  • value: organization-mechanism
  • activation: subject
  • patterns: organizational mechanism, strategy mechanism, capability, dynamic capabilities, organizational routines
• business.construct.managerial-implication
  • value: managerial-implication
  • activation: subject
  • patterns: managerial implication, business phenomenon, strategic management, competitive advantage

Method Lenses

Business should add method-level lenses that can be borrowed by other subjects:

method_lenses:
  business-positioning:
    resource: content/subjects/business/skills/business-journal-positioning-auditor.md
    activation: method_only
  qualitative-transparency:
    resource: content/subjects/business/overlays/skills/study-designer.md
    activation: method_only
  construct-level-fit:
    resource: content/subjects/business/overlays/skills/manuscript-architect.md
    activation: method_only

These lenses should not by themselves make business the primary subject.

Required Fixture Pack

Create tests/fixtures/subject_router_eval/business/ with at least these fixtures:

• clear_management_theory_case_study.json
  • clear positive
  • expected primary subject: business
  • expected method lenses: business-positioning, qualitative-transparency
• clear_marketing_platform_experiment.json
  • clear positive
  • expected primary subject: business
  • expected method lenses: business-positioning
• method_only_gioia_borrow.json
  • method-only borrow
  • expected primary subject remains auto
  • forbidden subject: business
  • expected method lens: qualitative-transparency
• mixed_finance_strategy_returns.json
  • mixed business-finance case
  • expected primary subject should stay with the dominant framing in the fixture; business may appear only when business theory contribution is explicit
• locked_economics_borrow_business_positioning.json
  • locked economics subject
  • expected decision: keep_locked
  • may borrow business positioning lens
• confirmed_business_journal_positioning.json
  • confirmed business subject
  • expected primary subject: business
• near_miss_small_business_plan.json
  • near miss
  • forbidden subject: business
• near_miss_consulting_market_analysis.json
  • near miss
  • forbidden subject: business
• near_miss_project_management_workflow.json
  • near miss
  • forbidden subject: business
• near_miss_teaching_case_assignment.json
  • near miss
  • forbidden subject: business

Each fixture must use subject_under_test: "business" and include tags that support the eval-ready gate:

• clear_positive,
• method_only_borrow,
• near_miss.

Expected Gate Behavior

After this slice:

business --gate eval-ready:
  activation_status: eval_ready
  eligible_for_eval_ready: true
  eligible_for_runtime_enabled: false
  blocking_failures: []

business --gate runtime-enabled:
  activation_status: eval_ready
  eligible_for_runtime_enabled: false
  blocking_failures:
    - activation_status is eval_ready

Accounting, finance, and economics runtime-enabled gates must remain green. Political economy, geoeconomics, and economics-accounting must remain deferred candidates and keep the onboarding blockers.

Expected Tests

Add or update tests in:

• tests/test_subject_contracts.py
  • business is eval_ready,
  • business has non-empty signal groups in all four dimensions,
  • business has method lenses,
  • business evaluation_pack points to the business fixture pack,
  • remaining deferred subjects still have blank evaluation_pack.
• tests/test_subject_router_eval.py
  • business fixture inventory includes required clear, method-only, mixed, locked, confirmed, and near-miss cases,
  • business eval-ready gate passes the real fixture pack,
  • business runtime-enabled gate fails on activation_status is eval_ready,
  • other deferred subjects still fail closed with onboarding blockers.
• tests/test_subject_refinement.py
  • clear business evidence can be measured under evaluation_subjects,
  • method-only business evidence borrows a lens without suggesting business,
  • runtime default does not activate business while business is only eval-ready.

Verification

Required focused tests:

uv run python -m pytest \
  tests/test_subject_contracts.py \
  tests/test_subject_router_eval.py \
  tests/test_subject_refinement.py \
  -q

Required gate checks:

uv run python tooling/scripts/evaluate_subject_router.py \
  --subject business \
  --gate eval-ready \
  --json

uv run python tooling/scripts/evaluate_subject_router.py \
  --subject business \
  --gate runtime-enabled \
  --json

uv run python tooling/scripts/evaluate_subject_router.py \
  --subject accounting \
  --gate runtime-enabled \
  --json

uv run python tooling/scripts/evaluate_subject_router.py \
  --subject finance \
  --gate runtime-enabled \
  --json

uv run python tooling/scripts/evaluate_subject_router.py \
  --subject economics \
  --gate runtime-enabled \
  --json

Required full eval:

uv run python tooling/scripts/evaluate_subject_router.py --json

Expected:

• Business eval-ready exits 0 with empty blockers.
• Business runtime-enabled exits non-zero because activation is still eval_ready.
• Accounting, finance, and economics runtime-enabled gates exit 0.
• Full eval metrics remain at required thresholds with no new near-miss false positives.

Final hygiene:

git diff --check
git status --short

Rollback

Rollback should restore business to:

activation_status: candidate
signal_groups:
  method: []
  data_or_outcome: []
  venue: []
  theory_or_construct: []
method_lenses: {}
evaluation_pack: ""

Business fixtures and tests can remain as candidate planning artifacts only if the gate keeps failing closed; otherwise revert the fixture pack with the manifest change.

Next Step After This Slice

If business eval-ready passes and review finds no precision issues, prepare a separate business runtime promotion design. That later design must add at least one auto-mode method-only guard and rerun accounting, finance, economics, and business gates together before changing business to runtime_enabled.