parcel-zoning-research

name: parcel-zoning-research version: "1.0" description: "Extracts and normalizes parcel, zoning, Official Plan, and overlay facts from a previously discovered Ontario source bundle. This skill should be used when source_bundle.json exists and the pipeline needs a source-backed normalized_data.json for downstream analysis." source: "Toronto Zoning By-law 569-2013, Toronto Official Plan, Ontario Parcel (LIO), TRCA" authority: "City of Toronto, Province of Ontario" data_as_of: "2026-03-07" pipeline_position: 2 pipeline_inputs: "source_bundle.json" pipeline_outputs: "normalized_data.json" pipeline_upstream: "source-discovery" pipeline_downstream: "buildability-analysis, precedent-research"

Parcel Zoning Research

Purpose

Convert a discovered source inventory into one normalized, source-backed parcel record. This skill performs extraction and normalization. It does not render a feasibility verdict.

Default operating assumption:

• Toronto is the primary fully specified jurisdiction
• Other Ontario municipalities are allowed, but the output must clearly label Toronto-specific fallbacks and unresolved fields

Use when

• source_bundle.json already exists
• The project needs normalized parcel and policy facts
• A downstream skill needs normalized_data.json

Do not use when

• Official source URLs have not been identified yet
• The task is to compare precedents or assess entitlement risk
• The task is to write the final client-facing report

Inputs

{
  "project_id": "uuid",
  "source_bundle_path": "source_bundle.json",
  "address_input": "123 Main St, Toronto ON",
  "project_concept": {
    "project_type": "multiplex"
  }
}

Required:

• source_bundle_path
• One of address_input or a parcel identifier present in the source bundle

Outputs

Produce normalized_data.json.

{
  "project_id": "uuid",
  "normalized_timestamp": "ISO datetime",
  "parcel": {},
  "zoning": {},
  "official_plan": {},
  "overlays": [],
  "applications_context": {},
  "unresolved_fields": [],
  "facts": [],
  "inferences": [],
  "assumptions": [],
  "unknowns": [],
  "overall_data_confidence": 0.0
}

Use this contract for zoning so downstream skills read consistent paths:

{
  "zoning": {
    "zone_code": "RD",
    "zone_label": "Residential Detached",
    "bylaw_name": "Zoning By-law 569-2013",
    "is_hatched_area": false,
    "legacy_bylaw": null,
    "chapter_900_status": "unverified | not_applicable | confirmed",
    "active_amendment_context": [],
    "development_standards": {
      "max_height_m": 10,
      "max_lot_coverage_pct": 33,
      "fsi": null,
      "min_front_setback_m": 3,
      "min_rear_setback_m": 7.5,
      "min_side_setback_m": 0.9,
      "parking_min_spaces": 1,
      "permitted_uses": [],
      "sources": []
    }
  }
}

Reference Router

Load only the references relevant to the task:

Workflow

1. Resolve the parcel

• Use the best official parcel or address source in source_bundle.json
• Resolve address, coordinates, and parcel identifiers first
• If the address cannot be uniquely resolved, stop and return an ambiguity warning instead of guessing

For Toronto:

• Prefer municipal address and parcel sources
• Use geocoding only as a fallback, and label it as such

2. Extract zoning facts

• Determine the applicable zone code
• Check for hatched-area or legacy-bylaw conditions
• Extract only the standards needed for feasibility screening:
  • use permissions
  • height
  • lot coverage
  • FSI where applicable
  • setbacks
  • parking

If multiple official sources disagree:

1. Record both
2. State which source is treated as controlling
3. Lower confidence

3. Extract Official Plan context

• Resolve the land-use designation
• Check for secondary plans, site-specific policy areas, or similarly material overlays
• Map the designation to compatible and incompatible use categories only when the policy support is explicit

4. Extract overlays and application context

Check independently for:

• heritage
• conservation-authority regulation
• flood or hazard mapping
• active planning applications
• recent permits when available

For each overlay entry, record:

{
  "type": "Heritage | TRCA Regulated Area | Floodplain | Active ZBA | Active OPA",
  "applies": true,
  "severity": "low | medium | high | unknown",
  "source_url": "https://...",
  "notes": "Short explanation",
  "confidence": "confirmed | inferred | unknown"
}

5. Apply Ontario overrides carefully

If the project concept depends on provincial override logic, such as recent Ontario housing rules:

• cite the governing provincial source
• identify exactly which local standard is affected
• record the override as a separate fact, not as a silent replacement

6. Normalize for downstream use

• Populate all four reasoning buckets: facts, inferences, assumptions, unknowns
• Put every unresolved but material field into unresolved_fields
• Keep downstream field paths stable; do not invent ad hoc keys

Jurisdiction Routing

Toronto

Run the full workflow using Toronto-specific sources and references.

Other Ontario municipalities

Run the same output contract, but:

• replace Toronto-specific sources with municipality-specific official sources
• mark Toronto-only checks as not applicable
• surface gaps instead of inventing municipality-specific logic

Quality Bar

• Every material extracted value must retain a source URL
• Every unresolved material field must be explicit
• Use development_standards.* paths consistently for all numeric zoning values
• Do not flatten the zoning contract in one skill and nest it in another

Stop Conditions

Stop and return a partial artifact when:

• parcel resolution is ambiguous
• the zoning source is blocked or contradictory
• a legacy by-law controls but cannot be retrieved

Reference Files

External References