Problem
records/sec_edgar.py queries only efts.sec.gov/LATEST/search-index (full-text search): relevance-ranked, 2001+, keyword-driven. There is no company-centric, date-ordered filing view anywhere — CIK is parsed out of results but never drives a query.
Real-world failure (2026-08-11): records search "Uranium Energy Corp" --source sec --limit 60 returned 4 UEC rows out of 60 — newest 2010 — the rest being other companies' filings that mention the name. The user's actual question ('what has this company filed recently') is unanswerable with the current primitive. One direct call to data.sec.gov/submissions/CIK0001334933.json answered it completely.
Proposal
New subcommand: openfoia records filings <TICKER|CIK> [--since DATE] [--forms 8-K,424B5,4]
data.sec.gov/submissions/CIK##########.json (+ paged archive files) — complete, chronological, per-company- ticker→CIK via SEC's
company_tickers.json
- reuse the existing adapter scaffolding (base.py egress/error handling is good)
- while here:
records fetch supports only documentcloud — EDGAR primary-document fetch falls out of this endpoint naturally (accession → Archives URL)
This is the single highest-value primitive for anyone using openfoia on public companies.
Problem
records/sec_edgar.pyqueries onlyefts.sec.gov/LATEST/search-index(full-text search): relevance-ranked, 2001+, keyword-driven. There is no company-centric, date-ordered filing view anywhere — CIK is parsed out of results but never drives a query.Real-world failure (2026-08-11):
records search "Uranium Energy Corp" --source sec --limit 60returned 4 UEC rows out of 60 — newest 2010 — the rest being other companies' filings that mention the name. The user's actual question ('what has this company filed recently') is unanswerable with the current primitive. One direct call todata.sec.gov/submissions/CIK0001334933.jsonanswered it completely.Proposal
New subcommand:
openfoia records filings <TICKER|CIK> [--since DATE] [--forms 8-K,424B5,4]data.sec.gov/submissions/CIK##########.json(+ paged archive files) — complete, chronological, per-companycompany_tickers.jsonrecords fetchsupports only documentcloud — EDGAR primary-document fetch falls out of this endpoint naturally (accession → Archives URL)This is the single highest-value primitive for anyone using openfoia on public companies.