Filter on property facts
The sample uses location, asking price and bedrooms. It does not use personal characteristics to rank buyers or determine housing eligibility.
INDUSTRY PLAYGROUND / REAL ESTATE
WebMCP real estate demo.
Filter fictional listings by city, price and bedrooms, then build a shortlist.
Checking browser support…Browser setup
Real estate playground
FICTIONAL DATA
OPEN / HOUSE
Find a place worth a look.
Filter fictional listings by city, price and bedrooms, then build a shortlist.
3 resultsShared page state
18 Cedar Lane
Austin · 3 bedrooms · 1,820 sq ft
cedar$585,000
204 Lake View
Austin · 4 bedrooms · 2,460 sq ft
lake$745,000
7 Maple Court
Portland · 2 bedrooms · 1,430 sq ft
maple$510,000
TOOL INSPECTOR 5 tools
Filter the visible fictional property listings by city, maximum USD price and minimum bedrooms.
LAST RESULT
// Run a tool to inspect its result.
Run tool calls the local handler directly. Native agent calls appear as “Native WebMCP” in the trace.
Execution trace 0 calls
Newest first. Example workflows are deterministic scripts, not LLM conversations.
No calls yet.
WHAT THIS EXAMPLE TEACHES
Make repeated actions predictable
Shortlisting the same home twice returns the existing selection. This is useful when an agent retries an uncertain response.
COMMON QUESTIONS
Are these real properties for sale?
No. Addresses, listing IDs and prices are synthetic demo data.
Does shortlisting contact an agent?
No. It changes a list on this page only. There is no lead form or external message.
API references reviewed September 7, 2026. Chrome imperative API ↗