Threat activity → Decisions lists the decisions: the bans (and captchas) CrowdSec has made, and the ones people added. The Defend-O-Tron’s firewall blocks every address with an active ban.
The page lists decisions from three origins:
cscli in the terminal.Community blocklists are enforced too, but not listed here. The line at the top says how many addresses they block now, and when the community blocklist was last updated. Show categories opens the breakdown (see Community protection).

| Column | What it shows |
|---|---|
| Decision | The type and how long is left, for example ban · 11h 20m left |
| Address or range | What is blocked. Select it to see its alerts |
| Origin | crowdsec, cscli or console (see above) |
| Scenario | The scenario that made the decision, for one from an alert |
| Country, Network (AS) | Where the address is |
| Until | When the decision ends |
| Alert | The alert that made it. Select it to open the alert’s details |
| Actions | Unban, for administrators |
Some decisions cover a range, for example the whole /64 of an IPv6 address.
The page shows 50 decisions at a time, with Older and Newer. It updates itself while you look at the newest page. On an older page it says Decisions have changed, and Show the newest goes back to the top.
Administrators can ban an address by hand, for example one that attacks a service the device has no scenario for.
203.0.113.7) or a range (198.51.100.0/24).The ban reaches the firewall within seconds, and is shared with every other bouncer connected to the device (see CrowdSec API). It is listed with origin cscli, like a ban made on the command line.

The device checks two things before it bans:
defend-o-tron-trusted). Click Ban anyway only if you are sure. See Allowlists.
Administrators can end a ban that this device made (origin crowdsec, cscli or console):
The firewall drops the ban within seconds. Other decisions on the same address stay, including community blocklists. The page then says whether anything else still blocks the address, such as another decision or an entry in Blocked addresses on the Firewall page.

Banning and unbanning need full CrowdSec access, such as the Admin role. A Viewer can see the decisions but not change them.
Real traffic sometimes trips a scenario. A scanner from your ISP’s network monitoring, a customer behind shared NAT, or a misconfigured legitimate service can be banned by mistake. There are four ways to handle it.
The lasting fix when an address should never be banned. Add it to Trusted addresses on Detection → Firewall. The change applies at once, without a restart, and public addresses are copied to CrowdSec’s allowlist, so CrowdSec stops acting on them too. A trusted address never makes evidence again, whereas an unbanned one is banned again the next time its traffic trips a scenario. See Honeypot for the trusted list.
See Unban above. Community blocklists are not listed here, so you cannot unban those addresses on this page. Add the address to the trusted list instead.
If the device is enrolled in a (free) CrowdSec account:
The unban reaches the device within a couple of minutes.
Open Device → Terminal and use cscli:
sudo cscli decisions list # see what is banned now
sudo cscli decisions delete -i 203.0.113.7 # remove a single address
sudo cscli decisions delete -r 198.51.100.0/24 # remove a range
sudo cscli decisions delete --id 42 # remove by decision ID
Unbanning an address does not stop it being banned again: its next attack bans it again. If the same address keeps coming back, add it to the trusted list, or find out why its traffic looks like an attack.