Derived fields

Proposes the record fields this property lacks, measured from the scan and the aerial photo.

What it answers

• What could we add to this record?
• What does the scan know that CoStar doesn’t?

For the agent, the tool is described as: For one scan, list the fields its CoStar or listing record lacks that the data can supply: area by room type, ceiling heights, door and window counts, glazed area, amenities, parking, roof material and condition, building height, footprint, pool. Each with value, method, confidence, evidence, and whether the record already has it.

Who it is for

CoStar research lead, CoStar researcher, Insurance underwriter.

Inputs

Input	Type	Required	Default	Notes
scan	string	yes		Scan id. Filled from the selected building.
aerial	boolean		true	Read the aerial facts (roof, height, footprint, pool); cached lookups.
vision	boolean		false	Also read panoramas and count stalls from above (uses the shared vision model). Saved preference truth.deep.

An input that nothing fills (the selected building, a saved preference, a persona’s preset) becomes a question to the person. The agent asks only for those.

Data used

Layer	Id	Needed	If it is missing
Rooms	matterport.rooms	required	The verdict is unknown, with the reason.
Property record	costar.property	optional	The tool runs without it.
Aerial photo	vexcel.ortho	optional	The tool runs without it.

What it returns

A tool result: a verdict (pass, warn, fail or unknown), metrics, checks and findings, each with its source.

• Panel blocks: VerdictHeader, MetricRow, MiniChart, CheckList, EvidenceStrip, StepLog.
• Speed: slow. Read-only.

How it works

Derived fields: the record fields a property lacks that the scan and the aerial photo can supply.

Reliability

Sources this tool gave on the test home:

Where it came from	Layer	Method	Reliability	Note
Measured from the scan	matterport	geometry	high	Matterport room areas and the wall graph
Measured from the scan	matterport	geometry	high	6 of 6 rooms have a measured height
Measured from the scan	matterport	geometry	high	Openings in the wall graph
Looked up in the scan	matterport	lookup	medium	Matterport room labels (by furniture)
Looked up in the scan	matterport	lookup	high

The tool sets the label for each value. A model never does. The tools give a scan’s wall graph the layer matterport whichever device made the scan, so a phone scan’s geometry carries that label too.

Limits

• Needs: rooms. On a scan without it, the verdict is unknown and says why.
• It is slow: progress steps reach the workspace while it runs.
• A result is a planning aid, not an inspection.

On the made-up test home

This section was written by running the tool when the docs were built, on a hand-made RoomPlan export (no real place was scanned), with no sponsor keys and the network switched off.

Pass 15 fields the record lacks, 0 it already has.

Value	Result	Where it came from	Reliability
Measured interior area	862 ft²	Measured from the scan	high
Typical ceiling height	8 ft	Measured from the scan	high
Tallest ceiling	8 ft	Measured from the scan	high
Doors and openings	6	Measured from the scan	high
Area, bedroom	344 ft²	Measured from the scan	high
Area, living room	324 ft²	Measured from the scan	high
Area, kitchen	113 ft²	Measured from the scan	high
Area, bathroom	48 ft²	Measured from the scan	high
Area, hallway	32 ft²	Measured from the scan	high
Windows	5	Measured from the scan	high

5 more values are in the full result.

Check	Measured	Against	Status
Measured interior area	862		pass
Area, bedroom	344		pass
Area, living room	324		pass
Area, kitchen	113		pass
Area, bathroom	48		pass
Area, hallway	32		pass
Typical ceiling height	8		pass
Tallest ceiling	8		pass

7 more checks are in the full result.

Findings: 15 pass.

What it reported while it ran:

1. Derived 15 fields from the plan

Run it

threshold run derived_fields --scan <scan-id>
curl -X POST http://127.0.0.1:8787/api/tools/derived_fields/run -H 'content-type: application/json' -d '{...inputs}'

Over MCP the tool is named derived_fields. See MCP server and CLI.