Skip to content

Honeytokens

Every bait that ReD-Pot plants is a honeytoken with a unique id: a password, an SSH key, an .env secret, a database hash, a GPP password, a Kerberoastable account or a Word document. When one is used, ReD-Pot knows:

  • what was stolen (the token),
  • where from (the decoy and path where it was planted),
  • who used it (source IP, workstation and user when known),
  • where it was used (the decoy and protocol).

Example alert: "password planted on RS-FS-01 used on RS-WEB-01 by ACC-PC-07 (anna.k)".

What counts as a hit

Action Level
a bait file is read or downloaded L4 compromise
a bait account name is used with a wrong secret (or appears in an AS-REP-roast user list) L4 (honeytoken.attempt), never L5
a bait credential is used with the correct secret: SMB, RDP, SSH, FTP, LDAP bind, web portal login L5 honeytoken
a Kerberos TGT is requested for a bait account with its password, or the honey SPN is Kerberoasted L5 honeytoken
a bait Word document is opened and reports home L5 honeytoken

SMB and RDP accept only a correct bait password (NTLMv2/v1 verified against its NT hash). Wrong passwords are logon failures and feed the password-spraying detection.

Credentials stolen at one site

Remote sensors receive a hashed index of every honeytoken. When a new sensor enrolls, the others pull the new index, so a credential stolen from one decoy is recognised on every decoy, at every site.

Bait documents

The file server's Word baits (Salaries_2026.docx, Network_Overview.docx) call home to the manager's web UI port when they are opened in Word. The alert names the host that opened the document.

Web UI allow-list

Bait documents only report from addresses that are allowed to reach the web UI. With a restricted web UI allow-list, a document opened on another host (for example an attacker's own workstation) is not reported. If the console is behind your reverse proxy, see Reverse proxy for bait-document beacons.

Where to see hits

  • Overview: Honeytokens hit counts the number of different baits used in the time range.
  • Deception: per decoy, the baits planted on it and how often each was hit.
  • Incidents / Hosts: the L5 incident and the attack path with the bait document.

See Breadcrumb chain for how baits on one decoy point to the others.