Constraint Satisfaction Validator

Step-by-Step Execution

Step 1: Retrieve constraint framework from knowledge base

search_knowledge_base(
    query="rational preferences constraints feasibility action space completeness transitivity",
    k=3
)

What to extract from the result:

• Ch.6 §6.1 — Constraints on Rational Preferences: rational agent requires completeness (every outcome is comparable) and transitivity — infeasible outcomes must be excluded before preference ordering is applied
• Hard constraint principle: an action is only admissible if it lies within the feasible region of the outcome space

Step 2: Retrieve robustness / stress-test framing for capacity check

get_entity_by_number(number="14.3")

What to extract from the result:

• Ch.14 §14.3 — Robustness Analysis: evaluate action under plausible deviations — use on_hand + reserved (not just on_hand) as the conservative capacity baseline to avoid overcommitment
• Stress-test boundary: if capacity utilisation > 0.9 (endpoint unavailable fallback), treat as at-risk and return CONDITIONAL rather than FEASIBLE

Step 3: Fetch warehouse capacity utilisation

GET /api/v1/smart-query/topology/warehouses/{warehouse_id}/capacity-utilization

Fields to extract: used_capacity_units, total_capacity_units, utilization_rate, location_id

Compute remaining capacity:

remaining_capacity = total_capacity_units − used_capacity_units

Fallback: if endpoint unavailable → assume utilization_rate = 0.9 (conservative stress-test default per Ch.14 §14.3)

Step 4: Fetch inventory balance and quality status

GET /api/v1/smart-query/inventory/observed-snapshot
    ?product_id=<product_id>
    &location_id=<location_id>

Fields to extract: on_hand, reserved, quality_status, last_count_at

Also fetch product shelf-life if applicable:

GET /api/v1/smart-query/topology/locations/{location_id}

Fields to extract: capacity_units, type

Step 5: Fetch order line for allocation bounds

GET /api/v1/smart-query/orders/{order_id}

Fields to extract per line: qty_ordered, qty_allocated, qty_shipped

Step 6: Evaluate hard constraints

Using the feasibility principle from Step 1:

# Hard Constraint 1 — Capacity
capacity_ok = (on_hand + qty_proposed) <= total_capacity_units
violation_1 = None if capacity_ok else "CAPACITY_EXCEEDED"

# Hard Constraint 2 — Quality (outbound shipments only)
quality_ok  = quality_status in ('good', 'verified')
violation_2 = None if quality_ok else f"QUALITY_BLOCKED: {quality_status}"

# Hard Constraint 3 — Shelf Life (if shelf_life_days available)
days_remaining = (expiry_date - today).days   # if applicable
shelf_life_ok  = days_remaining >= 3          # minimum viable shelf life
violation_3    = None if shelf_life_ok else f"SHELF_LIFE_CRITICAL: {days_remaining}d"

# Hard Constraint 4 — Allocation Bounds
allocation_ok  = qty_allocated + qty_proposed <= qty_ordered
violation_4    = None if allocation_ok else "ALLOCATION_EXCEEDS_ORDER"

# Collect all violations
hard_violations = [v for v in [violation_1, violation_2, violation_3, violation_4] if v]

Step 7: Evaluate soft constraints and compute violation score

Soft constraints do not block the action but reduce its utility score:

soft_violation_score = 0.0

# Soft: capacity utilisation > 80% after action
if (on_hand + qty_proposed) / total_capacity_units > 0.8:
    soft_violation_score += 0.2

# Soft: quality is 'acceptable' (not ideal but not blocked)
if quality_status == 'acceptable':
    soft_violation_score += 0.1

# Soft: last_count_at > 30 days (stale physical count)
if staleness_days > 30:
    soft_violation_score += 0.1

Step 8: Determine feasibility status

if len(hard_violations) > 0:
    feasibility_status = INFEASIBLE
    blocking_reason    = hard_violations[0]   # primary blocker

elif soft_violation_score > 0:
    feasibility_status = CONDITIONAL