Estimate per-unit renovation cost with line-item breakdown.

Args: unit_sqft: Unit square footage (drives size category: <700 small, 700-950 medium, 950+ large) bedrooms: Number of bedrooms bathrooms: Number of bathrooms (bathroom costs multiply by this) scope_level: "light" ($3K-$5K) or "standard_value_add" ($12K-$18K) finish_tier: "basic" or "upgraded" (granite, backsplash, stainless) year_built: Property year built (triggers risk flags) property_class: "B" or "C" market: "dallas" or "birmingham" unit_id: Optional unit identifier

Property-level estimate rolling up unit costs + exterior CapEx.

Args: property_id: Property identifier total_units: Total units at the property units: List of unit dicts with keys: unit_id, sqft, bedrooms, bathrooms, scope_level, finish_tier year_built: Property year built property_class: "B" or "C" market: "dallas" or "birmingham" exterior_items: Exterior CapEx categories to include (default: all). Options: roof, parking, hvac, siding_paint, fencing_gates, landscaping, signage_lighting

Property-level estimate using the Property entity model.

Accepts a single property dict containing all property context and unit mix. This is the preferred API for Plat — replaces loose kwargs.

Args: property_data: Dict with keys: - property_id (str, required) - total_units (int, required) - unit_mix (list of dicts with: sqft, bedrooms, bathrooms, scope_level, finish_tier, count) - year_built (int, optional) - property_class ("B" or "C", optional) - market (str, optional) - building_type (str, optional) - exterior_items (list of str, optional)

Check whether a renovation plan clears the minimum ROI threshold.

Uses the HIGH cost estimate (conservative). Formula: (annual_rent_lift / total_cost_high) * 100. Default threshold is 15%.

Args: total_cost_high: High-end total renovation cost estimate current_monthly_rent: Current monthly rent target_monthly_rent: Expected post-renovation monthly rent threshold_pct: Minimum ROI percentage (default 15%)

Generate a detailed scope of work (SOW) for contractor bidding.

Produces a contractor-ready document with material specs, quantities, and quality standards.

Args: unit_sqft: Unit square footage bedrooms: Number of bedrooms bathrooms: Number of bathrooms scope_level: "light" or "standard_value_add" finish_tier: "basic" or "upgraded" property_address: Property street address unit_id: Unit number/identifier

Evaluate a contractor bid against the internal cost estimate.

Flags: inflated line items (30%+ above benchmark), vague lump sums, unrealistic timelines.

Args: contractor_name: Name of the contractor bid_line_items: List of dicts with "description" and "amount" keys bid_total: Total bid amount timeline_days: Proposed timeline in days (None if not specified) unit_sqft: Unit square footage for generating comparison estimate bedrooms: Number of bedrooms bathrooms: Number of bathrooms scope_level: "light" or "standard_value_add" finish_tier: "basic" or "upgraded" year_built: Property year built

Estimate unit cost, validate ROI, and produce underwriting-ready renovation program input.

Runs the full pipeline: estimate → ROI gate → bridge dict matching the multifamily-underwriting deal schema's renovation_program shape.

Returns a dict with:

• renovation_program: ready to splice into the deal JSON

• roi_result: full ROI gate result for audit trail

• ready_to_underwrite: True if ROI gate passed

If the ROI gate fails, returns the result with ready_to_underwrite=False and the roi_result containing path_to_pass guidance.

Args: unit_sqft: Unit square footage (drives size category) bedrooms: Number of bedrooms bathrooms: Number of bathrooms current_monthly_rent: Current monthly rent target_monthly_rent: Expected post-renovation monthly rent start_month: Renovation start month ("YYYY-MM") monthly_pace: Units renovated per month scope_level: "light" or "standard_value_add" finish_tier: "basic" or "upgraded" year_built: Property year built (triggers risk flags) property_class: "B" or "C" market: "dallas" or "birmingham" unit_id: Optional unit identifier downtime_days: Vacancy days per unit during renovation (default 21) threshold_pct: Minimum ROI percentage (default 15%)

Idempotent property registration.

Accepts a Property-shaped dict (and optional inline floor_plans). Mints property_id if missing or returns the existing row when external_alias matches.

Inline floor_plans are deduped by (sqft, bedrooms, bathrooms): an existing match is reused, otherwise a new FloorPlan is created. This means calling register_property twice with the same alias and a NEW floor plan in the payload correctly creates the new floor plan on the second call (the previous behavior silently dropped it).

Canonical entry point: estimate a ScopeRequest, persist (request, estimate).

Returns the ScopeEstimate as a dict, or a ValidationProblem dict on error.

Project deal + scenarios → per-scenario ScopeRequest(s), estimate each, project back to renovation_program dicts.

Returns {estimates, renovation_programs} index-aligned with scenarios_list. estimates[i] is a ScopeEstimate dict on success or a problem dict (error_type-bearing) when the scenario hits the not_implemented guard. Per-type dispatch lands in slice B PR 3.

Return all (scope_request, scope_estimate, actual_outcome) triples for a property as a list of dicts. Empty list if property has no history.