SEC EDGAR Bash

Tiny collection of bash scripts for downloading and querying public filings straight from the SEC Data API. The scripts writes raw JSON into json/ and/or imports it into DuckDB for further exploration.

---

Prerequisites

• bash+, curl, duckdb+
• export EDGAR_UA to a descriptive User‑Agent string (SEC requirement)

export EDGAR_UA="ACME Corp research bot (contact: data‑team@acme.com)"

---

Script overview

Script	What it fetches	Minimal usage
get-submissions.sh	Entire submissions feed (company‑wide filing history).	./get-submissions.sh <CIK>
get-company-concept.sh	Time‑series for a single taxonomy concept (e.g. Assets).	./get-company-concept.sh <CIK> <taxonomy> <concept>
get-company-facts.sh	Full Company Facts bundle (all concepts for the filer).	./get-company-facts.sh <CIK>
get-xbrl-frames.sh	XBRL frame (cross‑company slice for a concept/unit/period).	./get-xbrl-frames.sh <taxonomy> <concept> <unit> <frame>
get-all.sh	Convenience wrapper that runs every above script for a demo set of concepts, logs output, and creates a DuckDB schema.	./get-all.sh <CIK>

---

Quick start

# pull everything Apple Inc. has filed and a XBRL frames for all companies
EDGAR_UA="Aktagon Ltd. (contact@aktagon.com)" ./get-all.sh 320193

# inspect data
duckdb edgar.db
SELECT * FROM company_facts LIMIT 20;

Output is stored in:

• json/ – raw API responses
• json-schemas/ – JSON schemas
• db-schemas/ – database schemas
• logs/ – curl + DuckDB progress
• edgar.db – local DuckDB with imported tables

---

File map

.
├── utils.sh               # shared helpers (CIK padding, cache probe)
├── get-all.sh             # orchestrator
├── get-company-concept.sh # concept time‑series
├── get-company-facts.sh   # full facts blob
├── get-submissions.sh     # filing history
├── get-xbrl-frames.sh     # peer frame slice
└── json/                  # API downloads

---

Troubleshooting

• 403 Forbidden → check that you have set the EDGAR_UA env variable and keep request rate within rate limit (≤10 req/s).

Happy parsing!