US Census Bureau

DATA · DATA & ANALYTICS

Public census datasets, variables, and geographies queried by year.

Connect US Census Bureau with Fabriq Developer

Your user pastes their US Census Bureau credential on a page Agentic Fabriq hosts, so it never reaches your backend and never becomes one shared key across all of your users. Same call as OAuth, one more argument.

from af_sdk import AgenticFabriq

af = AgenticFabriq()                      # reads AF_APP_ID / AF_APP_SECRET

url = await af.initiate_connection(request, provider="censusbureau", method="api_key")
return redirect(url)                      # an AF-hosted page, not yours

async with await af.for_user(request) as session:
    tools = await session.get_tools()     # only what this user granted

Provider names are surfaces, not vendors, and the authoritative list is the capabilities endpoint rather than any page including this one — an unknown name returns 400 with the valid list. Fabriq Developer is free to start and needs no provider paperwork. Start free or read the quickstart.

Connecting US Census Bureau in practice

A Census Data API query is capped at 50 variables and a key is required on every data request, so a wide table cannot be fetched by listing its columns — the group() wildcard is the documented way round it, and one group() call returns all 912 variables of table S0101. Responses are a two-dimensional JSON array whose first row is the column names, so there are no object keys for an agent to bind to.

Rate limits

Up to 50 variables in a single API query, separated by commas.
An API key must be used with all data queries made to the Census Data API, passed as a key= parameter in the URL.

Worth knowing

group(TABLE) returns every variable in a table even when the table has more than 50, which is the only way to read a wide table in one call.
Some datasets have required variables, and leaving one out returns no results rather than an error naming the missing parameter.
The vintage and dataset are path segments — /data/2019/acs/acs1 — so moving to a new year is a new URL, and a hard-coded path silently keeps serving last year’s data.
Because the response is a positional array rather than named objects, a variable added or reordered upstream shifts every column an agent reads.
The key travels in the query string, so it lands in any access log or proxy trace along the way.

Checked against Census Data API user guide (2026-09-29), Census Data API user guide (PDF), Query Limits (2026-09-29), Census API discovery tool (2026-09-29)

Acts as the person, not as itself

Each user connects their own account. Every call carries both identities — the agent and the person it is acting for — so the agent can never reach past what that individual can already do.

Credentials never touch the agent

Tokens live in the vault and attach server-side at call time. The agent holds a session, not a secret, and revoking access does not mean rotating a key.

Every call on the record

Who asked, which agent acted, which action ran, and the verdict that let it through — one audit trail across every integration, not one per vendor.

What an agent can do

Each action is granted on its own. An agent allowed to read is not thereby allowed to write, and the scope beside each row is what the acting user must have connected for it to run at all.

censusbureau_get_data_by_year_by_datasetREAD

Query a Census dataset: select up to 50 variables with `get` and narrow with geography (`for`/`in`) and other predicates. The answer is a JSON array of arrays with the header row first, e.g. [["NAME","state"],["Alabama","01"]] -- not objects. via GET /data/{year}/{dataset}

api

censusbureau_get_data_by_year_by_dataset_examples_jsonREAD

Example queries for a dataset: the available geographic levels and predicates, so data queries do not have to be built from scratch. via GET /data/{year}/{dataset}/examples.json

api

censusbureau_get_data_by_year_by_dataset_geography_jsonREAD

Geography list for a dataset: the geographic levels usable in the data query's `for`/`in` predicates. via GET /data/{year}/{dataset}/geography.json

api

censusbureau_get_data_by_year_by_dataset_groups_jsonREAD

Table groups for a dataset, usable with the data query's group([table ID]) get function. Filtered, compacted and paged at the tool layer (`query`/`limit`/`offset`): the provider serves the document whole (406 KB for 2021 acs/acs1). via GET /data/{year}/{dataset}/groups.json

api

censusbureau_get_data_by_year_by_dataset_sorts_jsonREAD

Sort list for a dataset ('SortList' link in the dataset discovery table). via GET /data/{year}/{dataset}/sorts.json

api

censusbureau_get_data_by_year_by_dataset_tags_jsonREAD

Tags list for a dataset. via GET /data/{year}/{dataset}/tags.json

api

censusbureau_get_data_by_year_by_dataset_variables_jsonREAD

Variables available in a dataset, including which are required. Filtered, compacted and paged at the tool layer (`query`/`limit`/`offset`): the provider serves the document whole and it reaches 13.5 MB (2021 acs/acs1, measured 2026-09-06). via GET /data/{year}/{dataset}/variables.json

api

censusbureau_get_data_jsonREAD

The master catalog of ~1,500 Census datasets (Project Open Data dcat:Catalog). Filtered, compacted and paged at the tool layer (`query`/`vintage`/`limit`/`offset`): the provider serves the catalog whole (5.2 MB, measured 2026-09-06). Each entry carries the /data/{year}/{dataset} coordinates for the other tools. via GET /data.json

api