Bondi & Bronte hunt
workspace7 Knox St, Clovelly |
14 Imperial Ave, Bondi |
12 Hewlett St, Bronte |
|
|---|---|---|---|
| Guide | $3.30m | $3.45m | $3.60m |
| Comp fair □ | $3.48m | $3.44m | $3.28m |
| vs fair | ~5% under | at fair | 8–13% over |
| $/m² land | $7.6k | $9.1k | $8.7k |
| Beds/baths/car | 3·2·1 | 3·1·1 | 3·2·1 |
| Land | 434m² | 372m² | 412m² |
| Station walk | 7 min | 9 min | 8 min |
| School decile | 10 | 9 | 10 |
| Confidence | High (7) | Med (4) | High (6) |
| Agent's read | Better buy Under fair on the biggest block, closest to a station, top school. |
Fair, but thin comps — small land for the $/m². | Over. Nice street, optimistic guide. |
Command canvas
Core interaction
No chat thread. A scrolling workspace of artifacts — the things you're working: a compare matrix, appraisals, a suburb profile. A command bar + chips (Appraise · Compare · Explain · What's new) materialise new artifacts onto the canvas. The latest answer surfaces at the top; older artifacts live below.
Why it fits a buyer's agent
The compare matrix is the centrepiece — the one thing portals can't compute, with comp-fair-value beside the guide and a plain-English "better buy + why". Frozen label column, value-coloured cells, an explicit Better buy verdict. KNOW-vs-RATE: computed cells are □ tinted; the agent's read is a separate, softer row. Every stat carries a provenance link.
Conservative ↔ inventive
The bridge between the chat spine (v1) and the spatial board (v3). You get a persistent, scannable working surface and a command grammar — but it still scrolls vertically rather than letting you arrange cards in 2-D space. Shippable on Lynx with a fixed component kit.
Real vs faked
REAL Clovelly/Bondi/Bronte, comp method, $/m², SEIFA decile, catchments, GTFS walk times, confidence-from-comp-count, agency-direct sourcing.
MOCK Addresses, prices, images (picsum), drag/share not wired.