SEC EDGAR full-text search — citation-ready filing matches for AI assistants

Run a keyless full-text search across SEC EDGAR filings via the official SEC EFTS API (https://efts.sec.gov/LATEST/search-index) and get back citation-ready results: company, CIK, form, filing date, accession number, document URL and relevance score for every matching filing.

Actor: apex-nexus~sec-edgar-full-text-search Store listing: not yet published — publication (FREE or PAY_PER_EVENT) is a director action and is gated by Apify's 5-publications-per-24h limit. This README must not be changed to claim a store listing until the director flips it.

What it is for

AI-assistant builders, legal researchers and compliance analysts who need to find what companies said in EDGAR filings (full text, not just the recent submissions metadata that most SEC scrapers return).

The Actor:

1. Calls the public SEC EFTS full-text search endpoint (https://efts.sec.gov/LATEST/search-index) with your query.
2. Parses every hit into a citation-ready dataset item: company, CIK, form, filing date, accession number, document URL and relevance score.
3. Caps output at maxItems (default 25, max 100) so a default run finishes in seconds.

Typical workflows:

• "What did companies say about X?" Search for a phrase (default: "material weakness") and get back every matching filing with a direct document URL.
• Company research. Restrict to a form type (forms=10-K) and scan filings for a term of interest.
• Verification. Every item carries accession, cik and documentUrl, so you can open the source SEC Archives document and confirm the match.

What this is NOT

This is explicitly not a tracker: no watchlist state, no delta alerts, no checkpoints, nothing remembered between runs. Each run is stateless and independent. It is distinct from N1 (SEC EDGAR filing extractor — returns citation-ready text chunks from a filing you bring), N2 (Form 4/3/5 XML to transaction table), N3 (SEC Company Facts XBRL to flat rows) and N6 (SEC 10-K/10-Q financial-statement extractor): N2-fulltext searches across all EDGAR full text rather than processing one filing you supply. The Actor is free when there is nothing to return.

Data sources

Public SEC data only, via the official SEC EFTS endpoint — no credentials, no proxies, no API keys, no LLMs, no commercial inference:

• https://efts.sec.gov/LATEST/search-index?q=...&forms=... — full-text search

One request per run, so SEC rate guidance is respected trivially.

Search terms (Store discovery)

Users find this Actor on the Apify Store with terms such as:

• SEC EDGAR full-text search
• EDGAR full text search
• SEC filing full text search
• sec full text search agent
• EDGAR citation-ready results

The Store title and description carry the core terms ("SEC EDGAR full-text search", "SEC filing full text search", "EDGAR full text search", "sec full text search agent", "citation-ready filing matches"); this section lists them for reference when updating the listing metadata.

Input

See .actor/input_schema.json. query defaults to "material weakness".

Field	Type	Default	Description
query	string	"material weakness"	Full-text search query passed to the SEC EFTS API.
forms	string	—	Optional comma-separated SEC form types (e.g. "10-K" or "10-K,10-Q,8-K"). All forms when omitted.
maxItems	integer	25	Hard cap on dataset items per run (1–100).

Example input (default run):

{
  "query": "material weakness",
  "maxItems": 25
}

Example input (10-Ks only):

{
  "query": "going concern",
  "forms": "10-K",
  "maxItems": 10
}

Output

One dataset item per matching filing. Download the whole dataset as CSV or JSON from the Apify run. See .actor/output_schema.json and .actor/dataset_schema.json.

Field	Type	Description
query	string	The search query this run used.
company	string	Company name as published in the EFTS record.
cik	string	SEC CIK (zero-padded to 10 digits).
form	string	SEC form type of the matching filing.
filingDate	string	Date filed with EDGAR (YYYY-MM-DD).
accession	string	SEC accession number.
document	string	Primary document file name (parsed from the EFTS hit id).
documentUrl	string (uri)	Direct URL of the SEC Archives filing document.
score	number|null	EFTS relevance score (higher = better match).
excerpt	string	EFTS returns scores, not excerpts — open documentUrl to locate the match.
retrievedAt	string	UTC timestamp (ISO 8601).

Example output item:

{
  "query": "material weakness",
  "company": "ChromaDex Corp. (CDXC) (CIK 0001386570)",
  "cik": "0001386570",
  "form": "10-K/A",
  "filingDate": "2020-05-18",
  "accession": "0001654954-20-005725",
  "document": "cdxc10ka_12312019.htm",
  "documentUrl": "https://www.sec.gov/Archives/edgar/data/1386570/000165495420005725/cdxc10ka_12312019.htm",
  "score": 9.922052,
  "excerpt": "",
  "retrievedAt": "2026-11-03T14:00:00+00:00"
}

Limits

• One search request per run; results are capped at maxItems (default 25, max 100).
• Invalid or unresolvable input finishes with an empty dataset rather than crashing.
• Requests are keyless and paced trivially below SEC rate guidance.
• The EFTS search-index endpoint returns relevance scores, not highlighted excerpts; excerpt is intentionally empty — open documentUrl for the matching text.

Quality gate (live, verified)

Actor: apex-nexus~sec-edgar-full-text-search (private — publication is a director action).

Default run (input {"query": "material weakness", "maxItems": 25}): returns >0 dataset items in under 5 minutes (verified via tools/smoke_test.py against the live EFTS API).

Claims statement

All claims in this README are limited to what the product verifiably does: it calls the public SEC EFTS full-text search API and returns citation-ready items (company, CIK, form, filing date, accession number, document URL, relevance score) for matching filings. This README contains no revenue, user, or adoption claims, because none are verified.