Posted on Sep 4 AI-assisted
I maintain a group of small sites that archive daily public records and related cultural reference material. The hard part was not fetching JSON; it was deciding when data from multiple sources was complete enough to publish.
I built a read-only REST API that normalizes the archive and aggregates 25 daily-updated series. Disclosure: I maintain the API and the linked 181649 / 49VIP sites.
curl 'https://api.181649.com/api/v1/latest'
Enter fullscreen mode Exit fullscreen mode
The API is free and requires no signup or API key. It includes:
- a stable response envelope;
- OpenAPI 3.1 and CORS;
- paginated historical records;
- 30 requests per IP per minute;
- issue, date, number-count, and cross-site consistency checks;
- explicit provenance stating that the two public domains are mirrors of the same aggregation source.
For example, I reject the latest aggregate rather than returning a partial success when any of the 25 series is stale, has a mismatched issue/date, or duplicates another signature. Clients receive a 502 and can keep a previously validated cache.
Docs: https://api.181649.com/docs?utm_source=devto&utm_medium=article&utm_campaign=public_api_launch
OpenAPI: https://api.181649.com/openapi.json
GitHub examples: https://github.com/tengleisun1-stack/49-gallery-api
The intended uses are archival work, integration testing, data-cleaning exercises, automation, and descriptive statistics. It does not offer wagering, payments, prize claims, or outcome guarantees.
I would especially value feedback on the response schema, provenance fields, and whether downloadable yearly snapshots would be more useful than pagination alone.