datagolisted

# Public data research — data.go.kr as a content seed Find and collect **open APIs and file datasets** related to a topic on the Korea open-data portal (data.go.kr), **and record them as sourced research seeds**. Source data from government and public bodies is itself a primary source — for the same number, prefer the original dataset over a news article requoting it (it also lightens the cross-check burden: 1 primary source = settled, 2 articles = cross-checked). ## Tool map | Tool | Auth | Use | |---|---|---| | `datago_search` | none | find datasets by keyword (API and FILE at once) | | `datago_detail` | none | get the collection path and metadata (last-modified, license scope) | | `datago_file_download` | none | download the file original (CSV, etc.) — **the fastest route** | | `datago_file_fetch` | key + 활용신청 (usage application) | page through large file datasets row by row (odcloud) | | `datago_api_call` | key + 활용신청 (usage application) | call a standard open API (live and query-conditional data) | The auth key is injected as `DATA_GO_KR_API_KEY` (project `.claude/settings.local.json`). Even with a key, a **per-API 활용신청** (usage application on the portal — log in; most are auto-approved) has to come first — see §4 below. ## Procedure ### 1. Search — find the datasets Search a keyword with `datago_search` (the arguments match the other search tools — `query`, `limit`, `page`). Tips: - On the first search, omit type (API and FILE together) — read the totals to see