9:41
◆ Mission control

Find me an Eastern
Suburbs home

3 tasks running for you right now. Tap any to watch, steer, or stop.
The brief I'm working to
Tune ✎

Hunt 3-bed houses under $3.5m in Bondi, Bellevue Hill, Randwick, Coogee & Vaucluse — must walk to a station, decent school catchment, and not be overpriced vs comps. Buying to live in, ready in ~3 months.

Min 400m² land ≤900m to rail No flood-prone + add must-have
214
Scanned
26
Passed brief
8
Shortlisted
Appraising new Bellevue Hill listings Running
On 31 Wallis Pde — found 6 sold comps, computing $/m² now. computeCMA()
4 of 7 listings · 0:08⏸ Steer
Watching for price-guide changes Running
Re-checking 8 shortlisted homes every 15 min for guide cuts & auction-date moves. changeDetect()
continuous · last hit 07:50⏸ Steer
Build compare matrix · top 4 Queued
Waiting on the Bellevue Hill appraisal to finish, then ranks your 4 strongest by value-for-money.
starts after task 1Run now ›
Overnight market sweep Done
Scanned 214 agency-direct listings → shortlisted 8 → culled 30 as overpriced vs comps.
★ 8 shortlisted − 30 culled ⌕ 47 sources
14 Imperial Ave, Bondi · top find
~6% under $3.35m guide vs $3.52m fair-value
↳ traced to 6 sold comps · SEIFA decile 9 ›
+
Give me a new task
"Also keep an eye on Tamarama under $4m"

Design notes — Variant B · The Dispatch Board close ✕

Core interaction

The brief sits at the top as editable instructions — tap any underlined token to tune budget, suburbs or must-haves. Below it, a board of parallel tasks (Cursor/Codex "mission control"): each run shows live status, an honest tool trace, and a Steer / Run now / Stop control. You delegate work, then supervise N streams.

Why it fits a buyer's agent

Frames the agent as staff you direct, not a chatbot you query. The funnel (214 → 26 → 8) makes its judgement legible at a glance, and a finished run hands back a reviewable artifact (top find + verdict) rather than prose.

Provenance & KNOW-vs-RATE

Every computed dollar figure is tagged blue and traces to its comps ("↳ 6 sold comps · SEIFA decile 9"). Verdict chips (under/over) are computed; the run's narration is the agent's read. Tool names (computeCMA()) keep the work honest.

Conservative ↔ inventive

The middle take. More ambitious than the inbox (A) — parallel runs, editable brief, intervene mid-run — but still a calm board, not the full immersive theatre of Variant C.

Real vs faked

Real: suburbs, the funnel logic, the data sources, the steerable-brief concept. Faked: addresses, dollars, timers and progress are illustrative; images are picsum placeholders.