Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Adversaries may communicate using a protocol and port pairing that are typically not associated.
.claude/skills/cyberstrikeus-t1571-non-standard-port/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 56% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -37% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 49% | 0% |
| case-22 | ✗→✓ | ▲ Improved | -1% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -68% | 0% |
Adversaries may communicate using a protocol and port pairing that are typically not associated. For example, HTTPS over port 8088 or port 587 as opposed to the traditional port 443. Adversaries may make changes to the standard port used by a protocol to bypass filtering or muddle analysis/parsing of network data.
Adversaries may also make changes to victim systems to abuse non-standard ports. For example, Registry keys and other configuration settings can be used to modify protocol and port pairings.
Platforms: ESXi, Linux, macOS, Windows
The following tests are from Atomic Red Team and provide actionable ways to test this technique:
Testing uncommonly used port utilizing PowerShell. APT33 has been known to attempt telnet over port 8081. Upon execution, details about the successful port check will be displayed.
Supported Platforms: windows
powershellTest-NetConnection -ComputerName #{domain} -port #{port}
Testing uncommonly used port utilizing telnet.
Supported Platforms: linux, macos
bashecho quit | telnet #{domain} #{port} exit 0
Dependencies:
If Atomic Red Team tests are not applicable, manually verify the technique by:
Properly configure firewalls and proxies to limit outgoing traffic to only necessary ports for that particular network segment.
Network intrusion detection and prevention systems that use network signatures to identify traffic for specific adversary malware can be used to mitigate activity at the network level.
| Finding | Severity | Impact | | -------------------------------------- | -------- | ------------------- | | Non-Standard Port technique applicable | Low | Command And Control |
| CWE ID | Title | | ------- | ---------------------------------- | | CWE-300 | Channel Accessible by Non-Endpoint |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-03 | fail→pass | 27,337 | 30,928 | +13% | 1 | 1 | 0% | 3,605 | 5,608 | +56% | 0 | 0 | — |
case-04 | pass→pass | 16,762 | 12,289 | -27% | 1 | 1 | 0% | 2,199 | 2,218 | +1% | 0 | 0 | — |
case-01 | fail→fail | 23,275 | 29,221 | +26% | 1 | 1 | 0% | 4,013 | 4,610 | +15% | 0 | 0 | — |
case-02 | fail→fail | 21,052 | 38,187 | +81% | 1 | 1 | 0% | 3,331 | 6,608 | +98% | 0 | 0 | — |
case-05 | pass→pass | 23,345 | 17,096 | -27% | 1 | 1 | 0% | 3,008 | 2,961 | -2% | 0 | 0 | — |
case-06 | pass→pass | 15,691 | 6,301 | -60% | 1 | 1 | 0% | 1,892 | 1,940 | +3% | 0 | 0 | — |
case-07 | pass→pass | 20,341 | 7,671 | -62% | 1 | 1 | 0% | 3,115 | 1,448 | -54% | 0 | 0 | — |
case-08 | fail→fail | 15,365 | 10,489 | -32% | 1 | 1 | 0% | 1,716 | 2,926 | +71% | 0 | 0 | — |
case-09 | pass→pass | 3,944 | 5,337 | +35% | 1 | 1 | 0% | 784 | 1,474 | +88% | 0 | 0 | — |
case-10 | fail→pass | 17,622 | 6,817 | -61% | 1 | 1 | 0% | 1,957 | 1,240 | -37% | 0 | 0 | — |
case-11 | pass→pass | 6,929 | 6,820 | -2% | 1 | 1 | 0% | 1,259 | 1,320 | +5% | 0 | 0 | — |
case-12 | fail→pass | 11,982 | 2,070 | -83% | 1 | 1 | 0% | 930 | 1,383 | +49% | 0 | 0 | — |
case-13 | pass→pass | 12,229 | 11,635 | -5% | 1 | 1 | 0% | 892 | 1,435 | +61% | 0 | 0 | — |
case-22 | fail→pass | 12,648 | 7,885 | -38% | 1 | 1 | 0% | 1,432 | 1,416 | -1% | 0 | 0 | — |
case-14 | fail→pass | 23,841 | 11,178 | -53% | 1 | 1 | 0% | 3,856 | 1,248 | -68% | 0 | 0 | — |
case-15 | fail→pass | 6,511 | 3,661 | -44% | 1 | 1 | 0% | 1,350 | 1,289 | -5% | 0 | 0 | — |
case-16 | fail→pass | 16,497 | 6,893 | -58% | 1 | 1 | 0% | 1,552 | 1,237 | -20% | 0 | 0 | — |
case-17 | pass→pass | 7,700 | 8,942 | +16% | 1 | 1 | 0% | 340 | 1,256 | +269% | 0 | 0 | — |
case-23 | pass→pass | 20,074 | 8,681 | -57% | 1 | 1 | 0% | 2,283 | 1,647 | -28% | 0 | 0 | — |
case-18 | fail→pass | 25,033 | 7,881 | -69% | 1 | 1 | 0% | 1,107 | 1,459 | +32% | 0 | 0 | — |
case-19 | fail→pass | 23,258 | 7,025 | -70% | 1 | 1 | 0% | 3,338 | 1,249 | -63% | 0 | 0 | — |
case-20 | pass→pass | 16,009 | 7,670 | -52% | 1 | 1 | 0% | 1,814 | 1,409 | -22% | 0 | 0 | — |
case-21 | fail→pass | 8,653 | 8,831 | +2% | 1 | 1 | 0% | 1,282 | 1,240 | -3% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted, and 22 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +43 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.