Federal Register — Multi-Query Document Export

Export regulatory publication records for recurring policy and industry research from the documented Federal Register API. Search by terms, agencies, document type, inclusive publication dates and optional CFR title/part. Each unique document includes sanitized abstract/action, agency identities, publication/effective/comment dates, citations, docket/RIN/CFR references and official PDF links.

Price: $0.0015 per unique delivered document. No start, query, duplicate, failure or empty-search charge. An optional Apify maximum charge reduces the global delivered-document cap. This independent Actor is not affiliated with NARA, OFR or any government agency.

Input

searchTerms: 0–5 native full-text terms, defaults to emissions. Empty means filter-only browsing and requires agency, type, an explicit date bound or CFR filter; it has no fallback search term. agencies: up to5 official slugs (empty=all); documentTypes: RULE, PRORULE, NOTICE, PRESDOCU (empty=all). publicationFrom and publicationTo accept optional inclusive YYYY-MM-DD dates; empty upper date resolves to UTC today and future dates are rejected. cfrTitle:0 disables,1–50 selects a title; cfrPart optional numeric part requires title. maxDocuments:1–500 global unique delivered cap; default20. Source order is newest publication first. These native searches do not guarantee topic relevance beyond the source index.

Example: {"searchTerms":["emissions"],"agencies":["environmental-protection-agency"],"documentTypes":["RULE"],"publicationFrom":"2025-01-01","maxDocuments":20}.

Output

One federal_document row per unique document, with matchedQueries across visited pages. The schema defines every column. Null dates/abstract/action mean the source omitted a value. A CFR reference part is null when the source cites only a title/chapter; no part is invented. COVERAGE in key-value storage records each query total, visited pages, stop reason, unvisited searches, global limits and whether complete; SUMMARY reconciles delivered rows and request counts. Coverage is not billed or placed in the dataset. Agent use: export EPA emissions publications in a date window with citation/agency/PDF for a human policy review. Schedule runs and compare identities outside the Actor; it stores no cross-run history and sends no alerts.

Limits

5 searches,500 unique rows,page50,30 wire requests including robots/redirects,180seconds acquisition,300second cloud timeout,256MB. Native API caps searchable paging at2000 results per term. Global or event cap can stop a search and leave later terms unvisited; KVS coverage records this. Data can change while paging; source count changes are recorded and duplicates removed, without claiming immutable/exhaustive corpus snapshots. Acquisition completes before <=50-row charged SDK batches. SDK delivery is not a full-run transaction and has no rollback guarantee. Any required source failure, invalid body, auth/rate-limit/timeout fails before paid output; genuine all-zero matches also fail with coverage and no charge. No browser, proxy, API key or paid external service.

Publication fields are informational API renditions. Official PDF links support checking original publications. Source-reported dates are not legal-effect, compliance or currently-open-deadline judgments; verify subsequent amendments/corrections and agency instructions for decisions. No sanction/KYB screening, full-text downloads or legal interpretation.

Privacy

Only public publication/organization metadata and source links. Email/phone patterns are removed from title/abstract/action/identifiers. No contacts, authors, comments, accounts, credentials, raw JSON, personal records or PDF text. Robots-disallowed private/UI paths are not requested. No official seals/logos.