Cribber data & AI access

Cribber compares agents and teams using 12 months of local listing activity, including active listings, homes with recorded sales, and other off-market listings. Each monthly leaderboard uses consistent geographic boundaries and crediting rules, with supporting listings you can inspect.

A home’s history does not end when it leaves the market. Cribber keeps qualifying past listings in the comparison, helping you see an agent’s broader listing record over time.

Public leaderboards and their downloads can be read with an ordinary HTTP GET request, without JavaScript, a Cribber account, an API key or an MCP connection.

Find a leaderboard

Choose a state in the HTML directory, then the exact city or town. Use both place name and state to distinguish namesakes.
Open the citywide leaderboard, or a listed neighborhood’s leaderboard where available. The city’s overview also links its published neighborhoods.
Read the leaderboard’s edition, selected metric and ranking table. Follow an agent or team name in the table to its individual public profile; follow supporting-listing links to inspect credited evidence.

HTML example: Massachusetts in the directory → Boston city overview → Back Bay, Boston agent leaderboard → an agent or team’s profile linked in that table.

South Carolina example: South Carolina in the directory → Mount Pleasant, SC citywide agent leaderboard. A city can have a citywide board without published neighborhood boards. Follow returned links rather than constructing slugs from names.

What the rankings measure

Credited listing / Homes Listed: Distinct qualifying homes credited to an agent or team during the reporting window. Current and past listings contribute under the same rules; co-listing agents each receive credit. Team or brokerage membership alone does not create credit.
Active, sold and inactive listings: Active: A home identified as actively listed at the observation or edition cutoff. Sold: A home with qualifying recorded-sale evidence matched to its listing episode. Inactive: An off-market listing without a qualifying recorded-sale match in the available evidence.
Rank: A position across the full eligible cohort for the selected metric, highest first. Equal metric values share a competition rank; tied positions can span pages. Filtering the table does not turn the visible subset into a new ranking.
Asking Volume: The combined selected observed asking prices of credited homes. A volume rank requires a usable asking price for every qualifying home. Missing prices do not remove homes from Homes Listed.
Agent or team: The canonical identity credited by the published evidence. Named teams can appear alongside individuals. Brokerage names provide affiliation context and do not award ranking credit.
Observed listing share: Credited homes divided by unique eligible homes in that leaderboard. Co-listing credits overlap, so agents’ shares can sum to more than 100%.
Geography and evidence: City boundaries and retained neighborhood assignments determine the comparison area. A property may count in both its city and neighborhood, once within each. Each edition identifies its qualifying homes, eligible agents and reporting dates, with supporting listing records linked for verification.

Read how Top Agents rankings work and each leaderboard’s own methodology and supporting records, such as Mount Pleasant’s published ranking methodology. Rankings and downloads use the existing published edition; this guide does not calculate or store rankings.

Monthly editions and data dates

An edition is a published monthly snapshot for one geography. The reporting window describes the listing activity included; the cutoff describes its data boundary; the publication timestamp records when the edition was published. Read those dates from the actual leaderboard rather than assuming a calendar-month window or using retrieval time as a freshness date.

Leaderboards capture verified monthly editions. Daily ingestion does not continuously rewrite a published edition. Historical editions preserve their original figures; a correction creates a new revision. Availability and the actual cutoff can differ between places. Use the page’s edition links or ?month=YYYY-MM to read an available historical edition.

JSON and plain-text directory reference

The JSON directory and the plain-text directory use the same public report catalog. They list published, search-enabled destinations. They contain navigation links, not a second set of rankings. The text version provides labeled links and paging instructions; the JSON version provides the fields below.

Filters and defaults

state: Optional two-letter US state code, accepted in either case; for example SC. Omit it for all states in the selected scope.
city: Optional exact citySlug returned by the directory. Use the returned neighborhoods link instead of guessing this value.
kind: city, neighborhood or all. Defaults to city, or neighborhood when a city filter is supplied. City includes city and town boards.
page: Page number from 1 to 10000; defaults to 1.
limit: Results per page from 1 to 100; defaults to 100.

Combine filters in the query string. Unknown parameters, repeated parameters and invalid values return HTTP 400. A temporary catalog failure returns HTTP 503; an empty successful result means no matching search-enabled destinations in that scope.

JSON response fields

schemaVersion, title, html: The directory schema version, description and canonical HTML directory URL.
scope: The selected state, city and kind; omitted state and city filters are represented as null.
states: State summaries within the catalog scope. Each has state, name, cities and neighborhoodCount, plus json and text links for city boards and an all JSON link including neighborhood boards.
total, page, limit, previous, next: The matching destination count and current page settings. Follow the exact previous/next URLs; a missing direction is null. Continue until next is null.
places: The current page’s destinations. Each includes name, kind, city, citySlug, state, and exact html, json and markdown links. City entries also include a neighborhoods directory link. That link can return an empty list.

Navigation examples

List South Carolina city and town boards as JSON; follow Mount Pleasant’s returned html or json link.
List Boston neighborhood boards as JSON; follow Back Bay’s returned links.
Read the same Boston neighborhood navigation as plain text.

Use the text or JSON directory if a retrieval tool cannot parse an XML sitemap or Markdown. The existing complete Markdown directory and search-enabled leaderboard sitemap remain available.

Read and cite the actual rankings

Open a returned html URL for the public ranking table. Its initial HTML includes the first page of agents, brokerages, ranks and listing counts, with ordinary pagination and profile links. Interactive sorting and search remain available on the website.

Use a returned json link for structured ranking data or markdown for a text table. These links request 25 agents with limit=25. Ranking exports have their own pagination object, including next, previous, totalAgents and fullDownload. Follow these links and preserve ties across pages. Omitting the export’s limit and page requests the complete edition; this differs from directory pagination.

Read selectedMetric, reportingMonth, window, cutoff, publishedAt, canonical and citation from the export. Agent rows include names, brokerages, credited listing counts, ranks, profile URLs and evidence links. Use the returned edition and pagination links to retain the metric and dates. publicationId in paged exports identifies the revision.

Cite the canonical leaderboard, the edition, selected metric and reporting window. Report tied first-place agents together. Follow the public page’s agent links for profiles and the edition’s supporting-listing links for provenance; use the matched recorded-sale evidence when reporting a home as sold.

For interactive structured queries, see Cribber’s public API and optional MCP documentation. For other public datasets, browse geographic Market Intelligence reports. Public content access and reuse follow Cribber’s AI access terms and crawler rules.