Similar rooms
Finds rooms by what they look like (‘a sunlit kitchen with an island’), with measured filters: over 12 m², in Houston, room for a wheelchair to turn.
What it answers
Section titled “What it answers”- A sunlit kitchen with an island, over 12 m², in Houston
- Bathrooms like this one with room to turn a wheelchair
For the agent, the tool is described as: Search a few hundred described rooms in Houston and Los Angeles by meaning, or find rooms like a given one. Filters on room type, city and measured area and turning circle in the same query. Returns rooms ranked by similarity, each with its description, measured area and a panorama to look at. Descriptions come from a vision model; sizes from the floor plan.
Who it is for
Section titled “Who it is for”CoStar research lead, CoStar researcher, Older adult or family, Renter checking a listing, Wheelchair user looking for a home.
Inputs
Section titled “Inputs”| Input | Type | Required | Default | Notes |
|---|---|---|---|---|
query |
string | e.g. ‘sunlit kitchen with an island and white cabinets’. Asked as: “Describe the room you are looking for” | ||
scan |
string | With room: find rooms like this one. Filled from the selected building. | ||
room |
string | Room id in scan. Filled from the selected building. |
||
label |
string | Room type: Kitchen, Bathroom, Bedroom, Living Room, … | ||
city |
string | Houston or Los Angeles. | ||
min_area_m2 |
number (m2) | Measured floor area at least. | ||
min_turn_m |
number (m) | Largest turning circle at least. | ||
words |
number | 0.3 |
Weight of exact word matches (BM25) against meaning. Range: at least 0, at most 1. | |
limit |
integer | 8 |
Range: at least 1, at most 30. |
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
Section titled “Data used”| Layer | Id | Needed | If it is missing |
|---|---|---|---|
| Room descriptions | threshold.facts.rooms |
required | The verdict is unknown, with the reason. |
What it returns
Section titled “What it returns”A tool result: a verdict (pass, warn, fail or unknown), metrics, checks and findings, each with its source.
- Panel blocks:
VerdictHeader,EvidenceStrip,CheckList,StepLog. - Speed: fast. Read-only.
How it works
Section titled “How it works”Similar rooms: search inside buildings by meaning. Each indexed room has a vision-model description of its panorama, embedded locally; StreamingDiskANN (pgvectorscale) finds the nearest, filtered by room type inside the index, and measured facts (area, turning circle, city) filter in the same query. BM25 word matching (pg_textsearch) can be blended in.
Reliability
Section titled “Reliability”Every metric and finding carries its layer, its method and a high, medium or low label, set by the tool for each value. See findings and reliability.
Limits
Section titled “Limits”- Needs: room descriptions. On a scan without it, the verdict is
unknownand says why. - A result is a planning aid, not an inspection.
On the made-up test home
Section titled “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.
Cannot answer here The facts database is not configured (TIGER_DATABASE_URL)
A phone scan has walls, doors, windows, rooms and objects, but no location, no photos and no property record.
Run it
Section titled “Run it”threshold run similar_rooms --scan <scan-id>curl -X POST http://127.0.0.1:8787/api/tools/similar_rooms/run -H 'content-type: application/json' -d '{...inputs}'Over MCP the tool is named similar_rooms. See MCP server and CLI.