Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Adversaries may manipulate application software prior to receipt by a final consumer for the purpose of data or system compromise.
.claude/skills/cyberstrikeus-t1474-003-compromise-software-supply-chain/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 23% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 12% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -68% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -30% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -55% | 0% |
> Sub-technique of: T1474
Adversaries may manipulate application software prior to receipt by a final consumer for the purpose of data or system compromise. Supply chain compromise of software can take place in a number of ways, including manipulation of the application source code, manipulation of the update/distribution mechanism for that software, or replacing compiled releases with a modified version.
Platforms: Android, iOS
Determine if the target mobile environment is susceptible to Compromise Software Supply Chain by examining the target platforms (Android, iOS).
Review whether mitigations for T1474.003 are in place. If defenses are absent or misconfigured, this technique may be exploitable.
Ensure Verified Boot is enabled on devices with that capability.
Security updates may contain patches that inhibit system software compromises.
| Finding | Severity | Impact | | ----------------------------------------------------- | -------- | -------------- | | Compromise Software Supply Chain technique applicable | Low | Initial Access |
| CWE ID | Title | | ------ | ------------------------- | | CWE-20 | Improper Input Validation |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 29,990 | 33,570 | +12% | 1 | 1 | 0% | 4,030 | 5,405 | +34% | 0 | 0 | — |
case-02 | fail→pass | 33,280 | 28,699 | -14% | 1 | 1 | 0% | 3,797 | 4,658 | +23% | 0 | 0 | — |
case-03 | fail→pass | 27,855 | 30,894 | +11% | 1 | 1 | 0% | 3,608 | 4,039 | +12% | 0 | 0 | — |
case-04 | pass→pass | 20,610 | 30,030 | +46% | 1 | 1 | 0% | 3,359 | 5,060 | +51% | 0 | 0 | — |
case-05 | pass→pass | 25,202 | 34,337 | +36% | 1 | 1 | 0% | 3,178 | 5,537 | +74% | 0 | 0 | — |
case-06 | pass→pass | 24,080 | 20,851 | -13% | 1 | 1 | 0% | 2,807 | 2,996 | +7% | 0 | 0 | — |
case-07 | pass→pass | 8,998 | 7,769 | -14% | 1 | 1 | 0% | 1,250 | 935 | -25% | 0 | 0 | — |
case-08 | fail→pass | 21,871 | 8,442 | -61% | 1 | 1 | 0% | 2,960 | 934 | -68% | 0 | 0 | — |
case-09 | fail→fail | 18,526 | 9,748 | -47% | 1 | 1 | 0% | 2,154 | 1,296 | -40% | 0 | 0 | — |
case-10 | fail→pass | 14,740 | 7,886 | -46% | 1 | 1 | 0% | 1,270 | 894 | -30% | 0 | 0 | — |
case-11 | pass→pass | 10,461 | 4,542 | -57% | 1 | 1 | 0% | 936 | 898 | -4% | 0 | 0 | — |
case-12 | fail→pass | 22,479 | 8,607 | -62% | 1 | 1 | 0% | 3,085 | 1,383 | -55% | 0 | 0 | — |
case-13 | fail→pass | 20,849 | 1,952 | -91% | 1 | 1 | 0% | 1,963 | 940 | -52% | 0 | 0 | — |
case-14 | fail→pass | 16,985 | 4,093 | -76% | 1 | 1 | 0% | 1,796 | 954 | -47% | 0 | 0 | — |
case-15 | fail→pass | 20,868 | 7,126 | -66% | 1 | 1 | 0% | 4,394 | 1,011 | -77% | 0 | 0 | — |
case-16 | fail→pass | 9,186 | 2,187 | -76% | 1 | 1 | 0% | 1,040 | 970 | -7% | 0 | 0 | — |
case-21 | pass→pass | 8,196 | 6,452 | -21% | 1 | 1 | 0% | 530 | 867 | +64% | 0 | 0 | — |
case-17 | fail→pass | 21,367 | 10,463 | -51% | 1 | 1 | 0% | 2,912 | 1,567 | -46% | 0 | 0 | — |
case-18 | fail→pass | 15,342 | 14,681 | -4% | 1 | 1 | 0% | 1,299 | 1,825 | +40% | 0 | 0 | — |
case-19 | fail→pass | 43,772 | 1,550 | -96% | 1 | 1 | 0% | 7,945 | 903 | -89% | 0 | 0 | — |
case-20 | fail→pass | 15,812 | 2,834 | -82% | 1 | 1 | 0% | 1,548 | 971 | -37% | 0 | 0 | — |
case-22 | pass→pass | 20,070 | 15,505 | -23% | 1 | 1 | 0% | 2,296 | 2,146 | -7% | 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 +59 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.