Bulk exports

Download the complete bills mirror as a stream designed for offline copies: gzip on the wire, one newline-delimited record at a time, and no client-side cursor loop.

GEThttps://api.archivist.dev/v1/exports/bills

Authentication

This is a programmatic API endpoint. Send an API key in the Authorization header on every request:

GET /v1/exports/bills···

curl --compressed https://api.archivist.dev/v1/exports/bills \
  -H "Authorization: Bearer $ARCHIVIST_KEY"

The export uses bearer-key authentication rather than browser sessions or cookies. Missing or invalid keys return 401; revoked or suspended access returns 403. See Errors & limits for the shared status reference.

Stream behavior

The endpoint has no query parameters and no cursor to follow. The server walks the full mirror internally in 500-row pages, then yields records through a pull-driven gzip pipeline so consumers can pipe the response to disk without buffering the export in memory.

What the client receives

200 OK. The success headers include Content-Type: application/x-ndjson and Content-Encoding: gzip.
The decoded body is newline-delimited JSON: every line is one BillItem object followed by \n.
Records are ordered by congress DESC, number DESC, id DESC.
Content-Disposition is attachment; filename="bills.ndjson.gz".

Download with curl

Use --compressed so curl advertises and transparently decodes the gzip response before writing plain NDJSON to disk. The same decoded stream can be piped directly to jq.

curl···

# Stream the full bills mirror; curl decodes Content-Encoding: gzip
curl --compressed --fail-with-body \
  -H "Authorization: Bearer $ARCHIVIST_KEY" \
  -o bills.ndjson \
  https://api.archivist.dev/v1/exports/bills

# Inspect the decoded NDJSON while piping
curl --compressed --fail-with-body \
  -H "Authorization: Bearer $ARCHIVIST_KEY" \
  https://api.archivist.dev/v1/exports/bills \
  | jq -c .

Consume with JavaScript

Standard Node and browser fetch implementations honor Content-Encoding: gzip before exposing response.body. Read chunks incrementally and split on newline boundaries; do not run the normally decoded body through a second gunzip.

export-bills.js···

const res = await fetch('https://api.archivist.dev/v1/exports/bills', {
  headers: { Authorization: `Bearer ${process.env.ARCHIVIST_KEY}` },
});
if (!res.ok) {
  const errorBody = await res.text();
  throw new Error(`Archivist ${res.status}: ${errorBody}`);
}
if (!res.body) throw new Error('Archivist returned no response body');

// Standard Node/browser fetch honors Content-Encoding: gzip before exposing response.body.
// Do not add a second gunzip. Use DecompressionStream('gzip') only when a lower-level
// client explicitly exposes raw compressed bytes.
const reader = res.body.getReader();
const decoder = new TextDecoder();
let buffer = '';

for (;;) {
  const { value, done } = await reader.read();
  if (done) break;
  buffer += decoder.decode(value, { stream: true });
  const lines = buffer.split('\n');
  buffer = lines.pop() ?? '';
  for (const line of lines) {
    if (!line.trim()) continue;
    const bill = JSON.parse(line);
    console.log(`${bill.id} — ${bill.officialTitle}`);
  }
}

buffer += decoder.decode();
if (buffer.trim()) {
  const bill = JSON.parse(buffer);
  console.log(`${bill.id} — ${bill.officialTitle}`);
}

Response schema

The response is a stream, not a JSON envelope. The table describes the success headers and the fields on each decoded BillItem line.

Response schema

Field	Type	Required	Description
status	200 OK	required	The export stream opened successfully.
Content-Type	application/x-ndjson	required	The decoded response body is newline-delimited JSON.
Content-Encoding	gzip	required	The body is gzip encoded on the wire. curl --compressed and standard fetch clients decode it before exposing the body.
Content-Disposition	attachment	required	Provides the download filename bills.ndjson.gz.
body	NDJSON stream	required	One BillItem object per line, with each line terminated by "\n".
body[].id	string	required	Stable bill id in the form {congress}-{type}-{number}.
body[].congress	integer	required	Congress number, for example 118 or 119.
body[].type	string	required	Bill type, for example hr, s, or sjres.
body[].number	integer	required	The bill number within its type and Congress.
body[].officialTitle	string	required	Official title from the bill mirror.
body[].introducedDate	string \| null	required	ISO 8601 calendar date, or null when unavailable.
body[].memberName	string \| null	required	Sponsor name, or null when no member is linked.
body[].memberBioguideId	string \| null	required	Sponsor Bioguide id, or null when unavailable.
body[].memberByCongressKey	string \| null	required	Congress-scoped sponsor key, or null when unavailable.
body[].committeeNames	string	required	Comma-separated committee names from the mirror.
body[].sourceUrl	string \| null	required	Canonical source URL, or null when unavailable.
body[].lastSyncedAt	string	required	ISO 8601 timestamp for the last mirror sync.

response.ndjson···

{"id":"118-hr-3076","congress":118,"type":"hr","number":3076,"officialTitle":"Civic Data Continuity Act","introducedDate":"2023-09-12","memberName":"Pelosi, Nancy","memberBioguideId":"P000197","memberByCongressKey":"118-h-P000197","committeeNames":"House Oversight,House Administration","sourceUrl":"https://www.congress.gov/bill/118th-congress/house-bill/3076","lastSyncedAt":"2026-04-22T03:11:08.412Z"}
{"id":"118-hr-3077","congress":118,"type":"hr","number":3077,"officialTitle":"…","introducedDate":null,"memberName":null,"memberBioguideId":null,"memberByCongressKey":null,"committeeNames":"","sourceUrl":null,"lastSyncedAt":"2026-04-22T03:11:08.418Z"}

Notes

The endpoint accepts no query parameters and exposes no cursor; one authenticated call covers the full mirror.
The export uses one rate-limit tick per call regardless of dataset size. A 429 response includes Retry-After with the next retry window.
Authentication and setup failures return JSON errors: 401 for missing or invalid bearer tokens, 403 for revoked or suspended access, and 500 for an unexpected setup failure.
Content-Disposition is attachment; filename="bills.ndjson.gz". The filename describes the wire encoding; curl --compressed and native fetch expose decoded NDJSON.
Internally the handler reads 500-row cursor pages and yields each line on demand, ordered by congress DESC, number DESC, id DESC.
If a database or gzip failure occurs after the 200 response starts, the connection terminates; treat an incomplete NDJSON line or file as a failed export.