Observed arrival · 2026-11-16
Decoyra’s AI-attacker honeypot study
A honeypot research project tests whether attackers follow instructions designed for language models and publishes threat data, including IP block lists, IOCs, and Suricata rules.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- Security researchers and incident-response teams
- Worth noticing
- Its seven-day snapshot includes 1,630 SSH sessions following the same recorded sequence across 58 addresses and 15 decoys.
Field notes
The site says each test hit is tied to the address given the instruction; forwarded, replayed, and browser-induced requests are excluded, with no scoring heuristics. Its weekly snapshot also lists 12 exploited vulnerabilities, 20 campaigns, and 49 malware files by hash; one SSH sequence—from reconnaissance through miner installation and cleanup—appeared in 1,630 sessions across 58 addresses and 15 decoys.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
One card from the complete issue