CyberMax-tools/eventwren-sec-8k-events

Eventwren: SEC 8-K material events by Item

Need it fresh, filtered or via API? This free file is a snapshot (SEC 8-K events up to the last refresh), last updated 2026-09-25.

• Insidewell on Apify ($0.004 per insider transaction): pulls today's SEC Form 4 trades for your own watchlist, filtered by buy/sell and size, with cluster-buy alerts on a schedule.
• Get an email when this dataset updates: free, double opt-in, unsubscribe any time.

Information only, not investment advice.

Every current report (Form 8-K and 8-K/A) filed with the SEC over the latest 20 EDGAR business days, split into one row per reported Item: new material agreements (1.01), earnings releases (2.02), acquisitions closed (2.01), impairments (2.06), delisting notices (3.01), auditor changes (4.01), restatements (4.02), CEO/CFO/director departures and appointments (5.02), cybersecurity incidents (1.05), bankruptcies (1.03) and the rest. Each row has the ticker, CIK, SIC industry code, SEC acceptance timestamp and a link to the filing.

This build (see meta.json): 3,847 filings → 8,354 item events, filed 2026-08-25 to 2026-09-22; 3,293 filings map to a listed ticker. They include 787 filings with officer/director changes (5.02), 686 new material agreements (1.01), 269 earnings releases (2.02), 7 bankruptcy filings (1.03), 4 non-reliance/restatement notices (4.02) and 3 material cybersecurity incidents (1.05).

Why a table and not EDGAR search? EDGAR's full-text search and RSS feeds don't let you filter by Item across all companies, which is what event-driven research, alert bots and LLM agents need ("every company that changed its CFO this month", "all 4.02 restatements").

Use it

from datasets import load_dataset
ev = load_dataset("CyberMax-tools/eventwren-sec-8k-events", "events", split="train").to_pandas()

# Leadership changes at listed companies, newest first
print(ev[(ev["item"] == "5.02") & ev.ticker.notna()][["accepted_at", "ticker", "company_name", "filing_url"]].head(20))

# Material events only (drops 9.01 exhibits and the 7.01/8.01 catch-alls)
print(ev[ev.is_material].groupby("item_title").size().sort_values(ascending=False))

DuckDB: SELECT ticker, company_name, accepted_at FROM 'hf://datasets/CyberMax-tools/eventwren-sec-8k-events/data/events.parquet' WHERE item = '4.02'.

Configs and columns

events (default, one row per filing × Item): accession_number, form_type (8-K or 8-K/A), filed_at, accepted_at (SEC acceptance time, US Eastern), period_of_report (date of the earliest event reported), cik, company_name, ticker, sic, item (e.g. 5.02), item_title (official Item name), is_material (Items 1.01–1.03, 1.05, 2.01, 2.02, 2.05, 2.06, 3.01, 4.01, 4.02, 5.01, 5.02), other_items_in_filing, filing_url.

filings (one row per filing): the same identifiers plus co_registrants (other entities on the same filing, e.g. trusts), sec_office (SEC industry office), state_of_incorporation, business_state, items (;-separated), item_count, has_material_item, is_earnings (2.02), is_exec_change (5.02).

Items come from the SEC's own header of each filing (<ITEMS> tags), not from parsing the document text, so they are exactly what the filer declared. The filing text itself is not included; follow filing_url. Ticker is null for issuers without an exchange listing in SEC's ticker file (e.g. ABS trusts, private debt issuers). Rows are never guessed.

Refresh

A rolling window: each refresh replaces the files with the latest 20 EDGAR business days. meta.json has the build time, window and counts. Built with direct EDGAR downloads that follow SEC's fair-access rules (declared User-Agent, under 10 requests/second).

Licence

SEC EDGAR filings and indexes are US government public records, free of copyright restrictions; this compilation is released under CC0 1.0. Compiled by CyberMax. Not affiliated with or endorsed by the SEC. Not investment advice.

Also from CyberMax

• Insidewell: SEC Form 4 insider trading tracker: insider buys and sells for any watchlist of tickers, with cluster-buy flags, on demand or on a schedule. Pairs well with 8-K events (an insider buy right after a 5.02 change is a classic signal).
• Free dataset: SEC Form 4 insider transactions, latest 5 EDGAR days.
• Free dataset: Figurewell: SEC XBRL financial statements and key metrics, the numbers behind each 2.02 earnings release.
• Free dataset: Riskroll: 10-K/10-Q Risk Factors and MD&A text.
• More tools for AI agents and data teams: https://apify.com/cybermax

More free from CyberMax

• All free CyberMax demos & datasets
• Insidewell: SEC Form 4 Insider Transactions (latest EDGAR filings)
• Curvewire: US Treasury yield curve + BLS macro panel (official sources)
• Insidewell: SEC Insider Trades

Quickstart notebook

Load this dataset with pandas and see real outputs in the quickstart notebook.