securitylevel
Set the Cloudflare security level (medium, high, or Under Attack) for one hosted domain or for all of them
- Command:
/go/securitylevel - Category: Cloudflare
- Required arguments:
<domain|all>and<level> - Output: One raw Cloudflare API JSON response per domain
- Last updated: 2026-11-21
On this page: Overview · Quick start · Syntax · Arguments · Output · Examples · Troubleshooting · Details · See also
Overview
/go/securitylevel changes a domain's Security Level setting on Cloudflare through the Cloudflare API. The security level controls how aggressively Cloudflare challenges visitors it considers suspicious. Use it to:
- raise protection for one site, or for every site, while it is being attacked
- return to normal protection when the attack is over
It can target:
- a single hosted domain, or
- every hosted domain at once (
all)
The command only changes the setting on Cloudflare. It does not change anything on this server. Domains must already be in Cloudflare (see dns).
Quick start
# Challenge more visitors on one domain
/go/securitylevel example.com high
# Turn on Under Attack Mode for every hosted domain
/go/securitylevel all under_attack
# Return every hosted domain to normal
/go/securitylevel all medium
Syntax
/go/securitylevel <domain|all> <level>
The command re-runs itself with sudo when it is not started as root, and writes the call to /var/log/go/securitylevel.log.
Arguments
| Position | Argument | Required | Notes |
|---|---|---|---|
| 1 | <domain> or all |
Yes | A hosted domain exactly as it appears in /home, or all for every hosted domain. |
| 2 | <level> |
Yes | One of the values in Levels. Any other value, or no value, sets medium. |
Levels
Matching is case-sensitive.
| You type | Applied to Cloudflare | What it does |
|---|---|---|
medium |
medium |
Normal protection. Challenges visitors Cloudflare rates as a moderate threat. |
high |
high |
Stronger protection. Challenges more visitors. |
under_attack |
under_attack |
Under Attack Mode. Every visitor sees a Cloudflare challenge page before reaching the site. |
underattack |
under_attack |
Same as above. |
under attack |
under_attack |
Same as above. Quote it: "under attack". |
Anything else (low, High, a typo, nothing) |
medium |
No error is shown. |
[!WARNING] A typo or a missing level does not fail. It sets
medium, which can silently lower a site that is currently onhighorunder_attack. Always check the level printed after the domain name, for exampleexample.com (high)....
The Cloudflare levels essentially_off and low cannot be set with this command.
Output
For each domain, the command prints one progress line, then the raw JSON that Cloudflare returned. It does not read the JSON, so the command always ends normally, even when Cloudflare rejects the change. Check the JSON for "success": true.
Single domain
example.com (high)...
{"result":{"id":"security_level","value":"high","editable":true,"modified_on":"2026-11-21T01:23:00.000000Z"},"success":true,"errors":[],"messages":[]}
All domains
Each line starts with a counter, [position/total].
[1/2] example.com (under_attack)...
{"result":{"id":"security_level","value":"under_attack",...},"success":true,"errors":[],"messages":[]}
[2/2] example.org (under_attack)...
{"result":{"id":"security_level","value":"under_attack",...},"success":true,"errors":[],"messages":[]}
Nothing printed
For a single domain, the command prints nothing if /home/<domain>/www doesn't exist. This is not an error message; the domain was simply skipped.
Examples
# One domain, strong protection
/go/securitylevel example.com high
# Spaces in the level need quotes
/go/securitylevel example.com "under attack"
# Under Attack Mode on every site during an attack, then back to normal
/go/securitylevel all under_attack
/go/securitylevel all medium
# Confirm the change was accepted
/go/securitylevel example.com high | grep -c '"success":true'
# Review recent calls
tail /var/log/go/securitylevel.log
Troubleshooting
| Symptom | Likely cause | Fix |
|---|---|---|
| Nothing is printed for a domain | /home/<domain>/www doesn't exist, so the domain was skipped. |
Check the spelling against the folders in /home. |
Level shown is medium but you asked for something else |
The level was misspelled, mistyped in another case, or missing. | Use medium, high, or under_attack, in lowercase. |
"success":false with an authentication error |
The Cloudflare API token is missing, wrong, or expired. | Update the token in GoAdmio's Cloudflare settings (see cloudflare). |
"success":false with a permission error |
The token can't edit zone settings. | Give the token the Zone Settings: Edit permission for the zones. |
"success":false with an invalid zone error |
The domain has no zone ID, or it is not on Cloudflare. | Add the domain with dns, then run again. |
The [n/total] counter doesn't match |
The total counts folders with a dot in the name, but the loop processes every folder that has a www directory. |
Cosmetic only. Every domain with a www folder is still processed. |
| Visitors see a challenge page on every site | Under Attack Mode is on. | Run /go/securitylevel all medium. |
Details
What it does
- Re-runs itself with
sudoif it was not started as root. - Appends the call and a timestamp to
/var/log/go/securitylevel.log. - Turns the level argument into
medium,high, orunder_attack(see Levels). - Reads the Cloudflare API token from
/var/lib/sqlite/config.db. - For each target domain that has a
/home/<domain>/wwwfolder:- Reads its zone ID from the
Sitetable in/var/lib/sqlite/site.db. - Sends a
PATCHrequest to Cloudflare'szones/<zone-id>/settings/security_levelendpoint. - Prints the response.
- Reads its zone ID from the
flowchart TD
A["securitylevel domain level"] --> B["Log the call"]
B --> C["Normalize level: medium, high, or under_attack"]
C --> D{"domain = all?"}
D -- "Yes" --> E["Loop over every folder in /home"]
D -- "No" --> F["Use the one domain"]
E --> G{"/home/domain/www exists?"}
F --> G
G -- "No" --> H["Skip silently"]
G -- "Yes" --> I["Read zone ID, PATCH Cloudflare, print JSON"]
Things to know
- Only domains with a
wwwfolder are processed. Alias, parked, and removed domains are skipped. - Domains are processed one at a time. There is no pause between requests, and a failure on one domain does not stop the others.
- The result is not verified. The command never checks
successin the response, and it doesn't return a failure exit code. - The setting is not undone automatically. Under Attack Mode stays on until you set another level.
- The Cloudflare account ID is read from the database but not used. Only the API token is needed.
- Plan limits apply. Cloudflare may restrict some levels on some plans; the response will say so in
errors.
See also
| Command | Description |
|---|---|
underattack |
Enable Cloudflare Under Attack Mode for a domain |
fightmode |
Enable or disable Bot Fight Mode on Cloudflare |
waf |
Block IPs via Cloudflare WAF |
block |
Block an IP address in the server firewall and in Cloudflare |
white |
Whitelist an IP address on this server and in Cloudflare |
purge |
Delete the Cloudflare edge cache (purge everything) |
proxy |
Change Cloudflare's proxy setting |
dns |
Add hosted domains to Cloudflare and set up their DNS zone |
cloudflare |
Configure Cloudflare DNS records, SSL, caching, and security settings |