Mementote-Harvester
Mementote-Harvester gathers the public record of places: old photographs, sound recordings, public-domain books and newspapers, and place names. It reads them from archives that publish them for reuse, through the doors they open: data dumps, mirrors, bulk datasets and official APIs. What it gathers becomes a quiet layer of public history under places in the Mementote app, every piece shown with its maker, its licence and a link back to you.
How it identifies itself
Every request carries this User-Agent:
Mementote-Harvester/1.0 (+https://www.mementote.co/harvest; harvest@mementote.co)
How it behaves
- It reads and obeys robots.txt, by the name
Mementote-Harvesteror*, including Crawl-delay. - It asks one thing at a time per site, spaced out, and waits as long as a 429 or 503 answer asks before trying again. Then it moves on.
- It sends If-None-Match and If-Modified-Since, so a file that hasn’t changed isn’t sent twice, and it resumes a broken download rather than starting over.
- It never signs in, never passes a paywall or a bot check, and never uses proxies or other addresses to get around a limit.
- It takes only what your terms allow. Public domain and Creative Commons works without a non-commercial term can be shown and played; a non-commercial work is linked to, never copied; anything without a clear licence is left out.
- It never reads social platforms, and never anyone’s private content.
- When it asks an archive about a place, it asks about a coarse grid square, never about where a person is standing.
Opting out
To keep it off all or part of your site, add this to your robots.txt. It is read again before each run.
User-agent: Mementote-Harvester Disallow: /
To have what it already gathered from you removed, or to ask it to go slower, write to harvest@mementote.co. A person reads every message, and we remove what you ask.
Questions about the app itself: Support.