Status
Every source, every day.
A collector that breaks quietly is worse than none. Each run is recorded per source so a store that changed its page shows up here before anyone notices stale prices.
| Source | Kind | Last run | Read | 7-day success | Avg time |
|---|---|---|---|---|---|
Adafruit | html | okNov 12 | 4/4 | 100% of 7 | 3.5s |
Mercado Livre | api | not running yet | 0/0 | — | — |
SparkFun | html | okNov 12 | 3/3 | 100% of 7 | 30.4s |
Steam (Brazil) | api | okNov 12 | 10/10 | 100% of 7 | 0.4s |
Steam (US) | api | okNov 12 | 10/10 | 100% of 7 | 0.7s |
Xbox Store (Brazil) | html | okNov 12 | 11/11 | 100% of 7 | 12.3s |
Xbox Store (US) | html | okNov 12 | 11/11 | 100% of 7 | 12.8s |
How collection works
- One run a day, from a scheduled GitHub Actions job; no server stays up.
- Every request carries an identified user agent and waits a second before the next one.
- Sources are read the way they publish data: official API, structured data on product pages, or a product feed.
- A product missing from a page or feed is skipped, never recorded with a guessed price.
What is collected
- Public product data only: name, price, list price, availability.
- No personal data from any store. Alert emails are confirmed by link and removed by link.
- robots.txt and terms are checked before a source is added.
- The collector and this site are one open repository on GitHub.