Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Expert firmware analyst specializing in embedded systems, IoT security, and hardware reverse engineering. Masters firmware extraction, analysis, and vulnerability research for routers, IoT devices, automotive systems, and industrial controllers. Use PROACTIVELY for firmware security audits, IoT penetration testing, or embedded systems research.
.claude/skills/dokhacgiakhoa-firmware-analyst/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | -26% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -29% | 0% |
| case-20 | ✗→✓ | ▲ Improved | -6% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 216% | 0% |
| case-06 | ✓→✓ | ＝ Same ✓ | -36% | 0% |
wget http://vendor.com/firmware/update.bin
screen /dev/ttyUSB0 115200
dd if=/dev/mtd0 of=/tmp/firmware.bin
## 🧠 Knowledge Modules (Fractal Skills)
### 1. [Hardware Methods](./sub-skills/hardware-methods.md)
### 2. [Phase 1: Identification](./sub-skills/phase-1-identification.md)
### 3. [Phase 2: Extraction](./sub-skills/phase-2-extraction.md)
### 4. [Phase 3: File System Analysis](./sub-skills/phase-3-file-system-analysis.md)
### 5. [Phase 4: Binary Analysis](./sub-skills/phase-4-binary-analysis.md)
### 6. [Authentication Issues](./sub-skills/authentication-issues.md)
### 7. [Command Injection](./sub-skills/command-injection.md)
### 8. [Memory Corruption](./sub-skills/memory-corruption.md)
### 9. [Information Disclosure](./sub-skills/information-disclosure.md)
### 10. [Extraction Tools](./sub-skills/extraction-tools.md)
### 11. [Analysis Tools](./sub-skills/analysis-tools.md)
### 12. [Emulation](./sub-skills/emulation.md)
### 13. [Hardware Tools](./sub-skills/hardware-tools.md)
### 14. [QEMU User-Mode Emulation](./sub-skills/qemu-user-mode-emulation.md)
### 15. [Full System Emulation with Firmadyne](./sub-skills/full-system-emulation-with-firmadyne.md)
### 16. [Checklist](./sub-skills/checklist.md)
### 17. [Reporting Template](./sub-skills/reporting-template.md)
### 18. [Finding 1: Title](./sub-skills/finding-1-title.md)
### 19. [Appropriate Use](./sub-skills/appropriate-use.md)
### 20. [Never Assist With](./sub-skills/never-assist-with.md)| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 17,673 | 30,292 | +71% | 1 | 1 | 0% | 3,216 | 2,608 | -19% | 0 | 0 | — |
case-02 | fail→fail | 23,231 | 21,425 | -8% | 1 | 1 | 0% | 3,325 | 2,528 | -24% | 0 | 0 | — |
case-03 | pass→pass | 4,772 | 8,222 | +72% | 1 | 1 | 0% | 413 | 1,307 | +216% | 0 | 0 | — |
case-04 | fail→fail | 17,879 | 14,198 | -21% | 1 | 1 | 0% | 2,970 | 2,474 | -17% | 0 | 0 | — |
case-05 | fail→fail | 26,857 | 24,299 | -10% | 1 | 1 | 0% | 2,347 | 2,776 | +18% | 0 | 0 | — |
case-06 | pass→pass | 10,093 | 4,381 | -57% | 1 | 1 | 0% | 1,881 | 1,197 | -36% | 0 | 0 | — |
case-07 | pass→pass | 7,776 | 4,898 | -37% | 1 | 1 | 0% | 1,599 | 1,258 | -21% | 0 | 0 | — |
case-08 | pass→pass | 9,724 | 8,193 | -16% | 1 | 1 | 0% | 1,573 | 1,841 | +17% | 0 | 0 | — |
case-09 | pass→pass | 6,813 | 2,985 | -56% | 1 | 1 | 0% | 907 | 973 | +7% | 0 | 0 | — |
case-10 | fail→pass | 14,700 | 5,845 | -60% | 1 | 1 | 0% | 2,085 | 1,544 | -26% | 0 | 0 | — |
case-11 | fail→pass | 7,844 | 5,126 | -35% | 1 | 1 | 0% | 1,287 | 909 | -29% | 0 | 0 | — |
case-12 | pass→pass | 8,058 | 5,023 | -38% | 1 | 1 | 0% | 1,326 | 879 | -34% | 0 | 0 | — |
case-13 | pass→pass | 11,553 | 3,121 | -73% | 1 | 1 | 0% | 1,440 | 866 | -40% | 0 | 0 | — |
case-14 | pass→pass | 10,916 | 6,371 | -42% | 1 | 1 | 0% | 1,537 | 1,598 | +4% | 0 | 0 | — |
case-15 | pass→pass | 9,611 | 6,767 | -30% | 1 | 1 | 0% | 1,538 | 1,623 | +6% | 0 | 0 | — |
case-16 | pass→pass | 11,336 | 4,813 | -58% | 1 | 1 | 0% | 1,575 | 1,291 | -18% | 0 | 0 | — |
case-17 | pass→pass | 5,437 | 5,160 | -5% | 1 | 1 | 0% | 818 | 1,206 | +47% | 0 | 0 | — |
case-18 | pass→pass | 5,887 | 4,423 | -25% | 1 | 1 | 0% | 781 | 1,155 | +48% | 0 | 0 | — |
case-19 | pass→pass | 7,340 | 5,809 | -21% | 1 | 1 | 0% | 1,266 | 1,409 | +11% | 0 | 0 | — |
case-20 | fail→pass | 8,487 | 4,871 | -43% | 1 | 1 | 0% | 1,500 | 1,404 | -6% | 0 | 0 | — |
case-21 | pass→pass | 13,675 | 11,606 | -15% | 1 | 1 | 0% | 2,174 | 2,442 | +12% | 0 | 0 | — |
case-22 | pass→pass | 12,155 | 25,494 | +110% | 1 | 1 | 0% | 1,542 | 1,979 | +28% | 0 | 0 | — |
case-23 | pass→pass | 17,229 | 7,348 | -57% | 1 | 1 | 0% | 2,417 | 1,715 | -29% | 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. The headline lift of +13 percentage points is the difference between those two pass rates over the 23 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.