PasserieBot
If PasserieBot turned up in your access logs, this page explains what it was
doing on your site, how often it will come back, and how to stop it.
How to identify it
| User-Agent | PasserieBot/1.0 (+https://passerie.com/bot) |
|---|---|
| robots.txt token | passeriebot |
| Operated by | Passerie, a moderated collection of recipes from independent food blogs. |
It does not crawl your site
PasserieBot does not discover pages, follow links, or walk a site. It fetches one URL at a time, and only after a registered Passerie user has submitted that exact URL by hand. If it requested a page of yours, a person asked for it by name.
A complete visit is normally two requests: the recipe page itself, and the one lead photograph that page declares. There is no sitemap fetch, no link following, and no periodic re-crawl.
How to block it
Add this to your robots.txt:
User-agent: passeriebot
Disallow: /
We check robots.txt before every fetch and cache it for one hour, so a rule you
add takes effect within the hour. We follow
RFC 9309. If your
robots.txt cannot be retrieved — a 5xx, a timeout, a network failure — we treat
that as a full disallow and do not fetch, as the RFC directs. Where the standard leaves us a
choice, we take the more restrictive reading.
Prefer not to edit robots.txt?
Email bot@passerie.com and we will block your domain at our end. The same address reaches us for removing recipes that are already published — see below.
How often it requests
| Between two requests to your site | At least 10 seconds |
|---|---|
| Between any two requests we make, anywhere | At least 2 seconds |
Crawl-delay |
Honored. It only ever lengthens the gap — nothing shortens it. |
| Request timeout | 20 seconds |
| Largest response we accept | 12 MB |
| Redirects followed | Up to 10, then we give up |
A page that returns 403, 404, or any other 4xx ends the submission on the first attempt. We do not retry into a refusal. Timeouts, 429s, and 5xx responses are retried later, on the assumption that the site is briefly unwell rather than telling us no.
What we read, and what we keep
We read the structured recipe data your page already publishes for search engines — JSON-LD or microdata. From it we keep:
- Title, and the short description
- The ingredient list
- Total time and yield, where stated
- Your name or publication name, and your domain
- The canonical URL
- A copy of the single lead photograph
We do not keep your method
Passerie stores no cooking instructions. There is no instructions field in our parser and no instructions column in our database — the steps are the part of a recipe worth visiting your site for, and we deliberately have nowhere to put them.
We copy the lead photograph so the stream renders at a consistent size without hotlinking from your servers. If you would rather we did not, tell us and we will remove it.
What happens to it then
Every recipe is reviewed by a person before it appears anywhere, and every recipe on Passerie credits you and links back to your page. The product is built to help someone decide what to cook and then send them to you to cook it — there is no reader mode, no ad-stripped view, and nothing to cook from on Passerie itself.
Removing your recipes
Email bot@passerie.com. Tell us your domain and we will remove what is published and stop fetching from you. You do not need to give a reason, and we will not ask you to justify it.
Passerie is in development and not yet open to the public. PasserieBot may already appear in your logs — the ingest pipeline runs against submitted URLs during development — so this page exists now rather than at launch.