個人 GitHub 帳號下公開儲存庫的擁有者可以登入並驗證所有權,自行管理介紹,無需管理員核准。
登入並產生驗證檔案,在 30 分鐘內將其提交至此儲存庫的預設分支。其他分支或分叉儲存庫中的檔案無效。然後在下方檢查檔案,即可無需管理員核准直接管理專案。請保留檔案以維持管理權限。
如需登入,將前往登入頁面。登入後請回到此處,再次點擊所選的驗證按鈕。
在專案資料夾中執行 /nomorevibe verify,即可驗證所有權並自行管理介紹。
一分鐘完成驗證Curated hyper-local dataset of US help and support services — recovery meetings, shelters, food banks, veteran and LGBTQ+ services, crisis lines — resolved to the city level
翻譯準備中
取自儲存庫 README
正在載入完整 README。 以下為已儲存的摘錄。
A curated, source-anchored, hyper-local dataset of help and support services across the
United States — mutual-aid meetings (AA, NA, SMART, ACA and more), NAMI affiliates and
peer support, shelters and housing assistance, food banks and pantries, veteran services,
LGBTQ+ centers, crisis lines, free clinics, legal aid — resolvable down to the city and
town level, plus a static Astro site that makes it searchable by location and need.檢視完整 README ↗A curated, source-anchored, hyper-local dataset of help and support services across the United States — mutual-aid meetings (AA, NA, SMART, ACA and more), NAMI affiliates and peer support, shelters and housing assistance, food banks and pantries, veteran services, LGBTQ+ centers, crisis lines, free clinics, legal aid — resolvable down to the city and town level, plus a static Astro site that makes it searchable by location and need.
Live site: succurro.org · roughly 257,000 records — organizations, service locations, and recurring meetings — across 32,307 US places, every one citing its source and carrying the date it was last verified.
Succurro (Latin): to run to the aid of, to help.
Tagged releases are archived on Zenodo with a DOI; publish/export_release.py
builds the citable bundle (JSONL + an Open Referral HSDS 3.0 export + schema +
statistics + SHA256SUMS). See publish/DATASHEET.md for the datasheet
(Gebru et al. framework), docs/RELEASE-CHECKLIST.md for the release gate, and
CITATION.cff for how to cite.
Same family skeleton as JUDGMENT, TOCSIN, and AUSPEX:
sources/ — raw downloads and scrape caches. Gitignored, regenerable, never
redistributed.data/ — the committed dataset. One YAML file per record, sharded by state and
city. Deterministic: same sources + same curated inputs ⇒ identical YAML. Never
hand-edit generated records; fix the pipeline or the curated inputs and rebuild.pipeline/ — Python (stdlib + PyYAML) module CLIs: python3 -m pipeline.<name>.
One module per source, each with on-disk caching, polite throttling, and loud payload
validation.site/ — static Astro site reading ../data at build time. Search via Pagefind
plus a build-time JSON index for location/category faceting.Core rules inherited from the family:
sources: list pointing at
first-class source records (URL, archive URL, retrieval date).verified:
stamp; the validator flags staleness. A wrong phone number in a crisis directory is a
safety issue, not a cosmetic one.null.YYYY-MM-DD. Enumerated fields are kebab-case tokens
checked against data/taxonomy/.data/
meta.yaml dataset version, counts, coverage
taxonomy/services.yaml controlled service-category taxonomy
places/<state>.yaml place registry: every city/town, Census-derived
orgs/<state>/<slug>.yaml organizations
sites/<state>/<place>/<slug>.yaml physical locations
meetings/<state>/<place>/<slug>.yaml recurring meetings/groups
sources/<publisher>/<slug>.yaml first-class source records
docs/
data-model.md the schema reference
sources.md source registry (status: in-use / planned / tested)
roadmap.md phased build plan
schemas/succurro.schema.json JSON Schema conformance gate
pipeline/ ETL modules
make help # list targets
make places # build the national place registry (Census gazetteer)
make verify # validation gate (schema conformance + referential integrity)
make install-hooks # pre-commit gate on data/ changes
Code is MIT (LICENSE). The dataset is CC BY-NC 4.0, with rights layered per
source — see DATA-RIGHTS.md.