Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Adversaries may overwrite or corrupt the flash memory contents of system BIOS or other firmware in devices attached to a system in order to render them inoperable or unable to boot, thus denying th...
.claude/skills/cyberstrikeus-t1495-firmware-corruption/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 22% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 29% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 47% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -36% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 82% | 0% |
Adversaries may overwrite or corrupt the flash memory contents of system BIOS or other firmware in devices attached to a system in order to render them inoperable or unable to boot, thus denying the availability to use the devices and/or the system. Firmware is software that is loaded and executed from non-volatile memory on hardware devices in order to initialize and manage device functionality. These devices may include the motherboard, hard drive, or video cards.
In general, adversaries may manipulate, overwrite, or corrupt firmware in order to deny the use of the system or devices. For example, corruption of firmware responsible for loading the operating system for network devices may render the network devices inoperable. Depending on the device, this attack may also result in Data Destruction.
Platforms: Linux, macOS, Network Devices, Windows
> Note: No Atomic Red Team tests available for this technique. See Atomic Red Team GitHub for updates.
Patch the BIOS and other firmware as necessary to prevent successful use of known vulnerabilities.
Prevent adversary access to privileged accounts or access necessary to replace system firmware.
Check the integrity of the existing BIOS and device firmware to determine if it is vulnerable to modification.
| Finding | Severity | Impact | | ---------------------------------------- | -------- | ------ | | Firmware Corruption technique applicable | Low | Impact |
| CWE ID | Title | | ------- | --------------------------------- | | CWE-400 | Uncontrolled Resource Consumption |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 44,944 | 38,054 | -15% | 1 | 1 | 0% | 6,057 | 5,835 | -4% | 0 | 0 | — |
case-02 | fail→fail | 39,463 | 51,919 | +32% | 1 | 1 | 0% | 4,927 | 7,984 | +62% | 0 | 0 | — |
case-03 | fail→pass | 35,193 | 32,802 | -7% | 1 | 1 | 0% | 4,199 | 5,141 | +22% | 0 | 0 | — |
case-04 | fail→pass | 26,734 | 26,048 | -3% | 1 | 1 | 0% | 3,202 | 4,142 | +29% | 0 | 0 | — |
case-09 | fail→pass | 29,377 | 30,998 | +6% | 1 | 1 | 0% | 3,281 | 4,820 | +47% | 0 | 0 | — |
case-05 | pass→pass | 16,542 | 5,878 | -64% | 1 | 1 | 0% | 1,652 | 1,538 | -7% | 0 | 0 | — |
case-06 | fail→pass | 24,209 | 13,213 | -45% | 1 | 1 | 0% | 3,255 | 2,087 | -36% | 0 | 0 | — |
case-07 | fail→pass | 18,969 | 24,227 | +28% | 1 | 1 | 0% | 2,056 | 3,736 | +82% | 0 | 0 | — |
case-08 | pass→pass | 12,916 | 18,184 | +41% | 1 | 1 | 0% | 1,955 | 2,810 | +44% | 0 | 0 | — |
case-10 | fail→fail | 23,048 | 25,587 | +11% | 1 | 1 | 0% | 1,631 | 2,237 | +37% | 0 | 0 | — |
case-11 | pass→pass | 10,203 | 3,529 | -65% | 1 | 1 | 0% | 801 | 1,145 | +43% | 0 | 0 | — |
case-12 | pass→pass | 4,717 | 10,099 | +114% | 1 | 1 | 0% | 601 | 1,625 | +170% | 0 | 0 | — |
case-13 | pass→pass | 17,996 | 9,508 | -47% | 1 | 1 | 0% | 2,247 | 1,673 | -26% | 0 | 0 | — |
case-14 | fail→pass | 15,659 | 4,603 | -71% | 1 | 1 | 0% | 1,459 | 1,629 | +12% | 0 | 0 | — |
case-15 | pass→pass | 22,262 | 25,085 | +13% | 1 | 1 | 0% | 2,535 | 3,725 | +47% | 0 | 0 | — |
case-16 | fail→pass | 32,774 | 36,290 | +11% | 1 | 1 | 0% | 4,490 | 6,399 | +43% | 0 | 0 | — |
case-17 | fail→pass | 10,567 | 7,400 | -30% | 1 | 1 | 0% | 1,483 | 2,051 | +38% | 0 | 0 | — |
case-18 | pass→pass | 25,182 | 21,033 | -16% | 1 | 1 | 0% | 2,601 | 3,340 | +28% | 0 | 0 | — |
case-19 | pass→pass | 18,544 | 21,091 | +14% | 1 | 1 | 0% | 2,746 | 3,292 | +20% | 0 | 0 | — |
case-20 | pass→fail | 25,108 | 48,352 | +93% | 1 | 1 | 0% | 3,779 | 7,546 | +100% | 0 | 0 | — |
case-21 | fail→pass | 21,671 | 28,821 | +33% | 1 | 1 | 0% | 3,380 | 5,104 | +51% | 0 | 0 | — |
case-22 | pass→pass | 20,006 | 18,787 | -6% | 1 | 1 | 0% | 3,233 | 4,229 | +31% | 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. 22 cases were attempted. The headline lift of +36 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
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.