#anubis — Public Fediverse posts

https://fzakaria.com/2026/07/09/who-does-anubis-actually-stop

The answer I can very helpfully give this blog post is "Abusive Scrapers"

I'm not trying to block AI, that's a battle I am sure as shit not going to win.

But naive scrapers that absolutely demolish my website?

Works a treat.

And I'm sorry it's come to this because yeah, it fucking sucks. It's repulsive. But it's also necessary.

#Anubis #Blog #AI

Fediverse server behind Anubis

There's a lot of scrappers and bots trying to get access to fedi posts and media all the time.

My CDN balance was running out quickly due to bots downloading media over and over again.

Then I installed Anubis and things are better now. It isn't a silver bullet, as some bots have enough resources to beat the challenge (like the ones from Tencent), but it's better than nothing and I don't need to resort to Cloudflare's basket.

If you're interested in protecting your fedi server, here's a starting point: https://blogs.gayfr.social/barbapulpe/protect-mastodon-against-ai-bots-with-anubis

I know that many people are against using Anubis, but now there's lighter JS challenges (preact) or no JS at all (metarefresh). Soon there should be TLS fingerprinting and other things available too.

Anyway, this post is for those who see Anubis as an option. If you don't, there's still Fail2Ban, blocklists and the like. Be safe out there!

#MastoAdmin #FediAdmin #Anubis

I found Anubis to be incredibly effective at relieving my server from the endless harassment from AI scrapers, but the Anubis documentation was bad and wrong. So I wrote a new guide to configuring Anubis with nginx.

[ANUBIS] - Ransomware Victim: Interim HealthCare [Head office] - https://www.redpacketsecurity.com/anubis-ransomware-victim-interim-healthcare-head-office/

#anubis #dark_web #data_breach #OSINT #ransomware #threatintel #tor

New post from #Anubis : Interim Healthcare [Head Office]
More at : https://www.ransomlook.io/group/Anubis #Ransomware

✅️ Migrated another #WordPress site to #Hugo.

1️⃣ I only post photos with no text, so it was cumbersome to create single WordPress posts for each photo.

2️⃣ Hugo has a way to loop through a directory, so now I simply loop through a folder with 2000 photos and print <img> with loading lazy.

😨 My self-hosted #Gitea instance had a moment of panic as it couldn't handle a commit with 700 Mb in assets. But it's all good now.

🤖 I've also added #Anubis in front.

🔗 https://ricard.photos

[ANUBIS] - Ransomware Victim: Scholle IPN / SIG - https://www.redpacketsecurity.com/anubis-ransomware-victim-scholle-ipn-sig/

#anubis #dark_web #data_breach #OSINT #ransomware #threatintel #tor

New post from #Anubis : Scholle Ipn / Sig
More at : https://www.ransomlook.io/group/Anubis #Ransomware

😩 Had a hard time finding resources online to help me build a useful #Grafana dashboard for #Anubis

✅️ I've added protection to 3 sites

🤔 Not 100% happy, I will continue to play with this.

I've made #Anubis work for one of my sites as a test run with #Docker and a fine tuned #NGINX config:

🔗 https://manga.ricard.blog

What I'm having trouble with is setting the bot protection policy to be more "difficult" (higher setting than the default 2) without losing the default policies that come out of the box 🤔

🔗 https://anubis.techaro.lol/docs/admin/policies/

Investigating how to setup Anubis to block AI traffic to my server given my current setup (NGINX, with some static sites as well as Docker apps)

🔗 https://github.com/TecharoHQ/anubis

#Anubis #NGINX

[ANUBIS] - Ransomware Victim: Interim HealthCare - https://www.redpacketsecurity.com/anubis-ransomware-victim-interim-healthcare/

#anubis #dark_web #data_breach #OSINT #ransomware #threatintel #tor

@tusharhero a lot of smaller websites put up #Anubis these days but the message there is very apologetic so I give them a free pass. Normally, it works out if I enable it temporarily to run the Anubis challenge and then disable it again (I should probably set up a whitelist or something for this; just haven't got around to it yet)

𝗘𝘅𝗰𝗹𝘂𝘀𝗶𝘃𝗲: 𝟱𝟬𝟬 𝗛𝗼𝘀𝘁𝘀, 𝟭 𝗧𝗕 𝗮𝗻𝗱 𝗡𝗼 𝗡𝗲𝗴𝗼𝘁𝗶𝗮𝘁𝗶𝗼𝗻: 𝗔𝗻𝘂𝗯𝗶𝘀 𝗥𝗲𝘃𝗲𝗮𝗹𝘀 𝗜𝘁𝘀 𝗙𝗮𝗶𝗿𝗹𝗶𝗳𝗲 𝗔𝘁𝘁𝗮𝗰𝗸

A vulnerable VPN, weak passwords, one week to reach the highest privileges, and approximately 500 compromised hosts. Anubis tells SuspectFile how it allegedly attacked Fairlife, a wholly owned subsidiary of The Coca-Cola Company. The group claims to have exfiltrated approximately 1 TB of data, 546,573 files, and demanded $15 million. But above all, it claims that this time there was never a real negotiation.

https://www.suspectfile.com/500-hosts-1-tb-and-no-negotiation-anubis-reveals-its-fairlife-attack/

#Anubis #Coca_Cola #Data_Breach #Fairlife #Milk #Ransomware

Just moved away from clourflare tunnels to hosting my website back on 1984.hosting. I use Anubis to block ai scrappers and have setup a tor hidden service as a backup.

You can get to it at the normal domain: godfrey.online
And the Tor address: rengrm2kdmv3mwxkvizyupyo3bcol3fhe7dqzydttpgnom6c3yni43id.onion

NOTE: Floatplane and MeansTV RSS should still work and are excluded from Anubis.

#cloudflare #anubis #tor #indieweb

New post from #Anubis : Interim Healthcare
More at : https://www.ransomlook.io/group/Anubis #Ransomware

DE: Beobachten andere Community-Projekte derzeit ebenfalls ungewöhnlich starke Scraper- oder #DDoS artige Zugriffswellen auf öffentlichen Git-/ #Forgejo Instanzen oder anderen Diensten?

Wir bei #Friendica beobachten solche Probleme seit einigen Tagen immer wieder mit wechselnder Intensität. Heute kamen innerhalb von vier Stunden rund 122.000 Anfragen von über 100.000 weltweit verteilten IP-Adressen, mit 172 verschiedenen, offenbar gefälschten Browser-User-Agents und auffällig wechselnden Sprach-Headern. Betroffen waren vor allem tiefe Commit-, Source- und Media-Pfade. #Anubis hat den Großteil abgefangen.

EN: Are other community projects currently seeing unusually heavy scraper or DDoS-like traffic against public Git/Forgejo instances or other services?

We at Friendica have been experiencing this kind of activity intermittently for several days, with varying intensity. Today, within four hours, we received around 122,000 requests from more than 100,000 globally distributed IP addresses, using 172 different, apparently spoofed browser user agents and conspicuously varying language headers. The requests mainly targeted deep commit, source and media paths. Anubis absorbed most of it.

Interested in whether this is part of a broader pattern.

New post from #Anubis : Cleaver-Brooks
More at : https://www.ransomlook.io/group/Anubis #Ransomware

[ANUBIS] - Ransomware Victim: Cleaver-Brooks - https://www.redpacketsecurity.com/anubis-ransomware-victim-cleaver-brooks/

#anubis #dark_web #data_breach #OSINT #ransomware #threatintel #tor

To me it seems that this anti-bot check thing is just broken in this scenario and since the invidious instances use the same anti-bot stuff, this affects all of them, afaik.

The main issue is, that the "forbidden!" page doesn't have any link or so to try anew, and that the URL is changed so much so, that I can't quickly manipulate it back to the normal video URL _and_ that the back button doesn't work ... Silly JS things.

#anubis #gandalf

Spends a couple of minutes updating the servers' #Anubis container to the latest release before going on a coffee run. ☕

One of my repositories on my self-hosted forgejo instance is being hit by Anthropic crawlers. They have an infinite loop to download a tar archive of the repository leading to somewhat like 200GB of archives. My backup process struggled to send snapshots remotely on my 6 Mbps upload line. I've made the repository private. And now I'm considering putting Anubis in front of the service.

#forgejo #anubis #ai #anthropic #selfhosting #homelab

@mwichary @michaelgemar The #Shieldfont site has a section on this, it is basically doing #Anubis-style POW for screen readers.

Thanks to @j-g00da , you can now bless any url!

https://anubis.kut.as

Stay prayed up!

#anubis #unix_surrealism #japanese_jesus #stayprayedup

Initially I locked down my site with relatively strict #anubis settings. Now I'm trying to open it up a bit. I've switched to a customized robots.txt in an attempt to let the #google search tools in. But now I'm getting double spend errors with Google's crawling attempts. Error messages blame me, but cursory searches claim that it's the client that's misbehaving. Who is really to blame?