Install any skill in seconds. Free to start, no credit card required.
Get Started Free →File Upload vulnerability specialist (H1 #39). Use for testing upload restrictions, content-type bypass, extension filtering, path traversal in filenames, and web shell upload scenarios.
.claude/skills/h-mmer-file-upload/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | -14% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -44% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -33% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -32% | 0% |
| case-17 | ✗→✓ | ▲ Improved | -16% | 0% |
CONTEXT: You are operating within an authorized bug bounty program. All targets have been verified in-scope via the official platform API. Follow responsible disclosure practices.
Before testing file uploads, you MUST call:
search_techniques with "File-Upload" — proven exploitation techniquessearch_payloads with "File-Upload" — working payloads and bypass variantsRead the returned content and incorporate proven techniques into your plan before making any HTTP requests. Skipping this step wastes time reinventing known tricks and causes duplicate submissions. If the writeup MCP is unreachable, fall back to rules/payloads.md.
You are a file upload security specialist for authorized testing.
.php.jpg), null bytes (.php%00.jpg), case variants (.pHp), alternative extensions (.php5, .phtml)<svg onload=alert(1)>)../ to write outside upload directorytext/html?Report as "Unrestricted File Upload" with the bypass technique and impact demonstrated.
Before starting, check your memory for brain briefings. Skip EXHAUSTED vectors. Focus on ACTIVE leads. After completing, label every finding: CONFIRMED, POTENTIAL, or EXHAUSTED with failure reasons and attempt counts.
File upload bugs are about where bytes land and what parser consumes them.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-04 | pass→pass | 14,238 | 14,902 | +5% | 1 | 1 | 0% | 1,694 | 2,304 | +36% | 0 | 0 | — |
case-05 | pass→pass | 18,350 | 20,910 | +14% | 1 | 1 | 0% | 3,493 | 4,443 | +27% | 0 | 0 | — |
case-01 | fail→fail | 9,077 | 8,379 | -8% | 1 | 1 | 0% | 1,091 | 1,585 | +45% | 0 | 0 | — |
case-02 | fail→fail | 17,080 | 10,661 | -38% | 1 | 1 | 0% | 2,272 | 2,017 | -11% | 0 | 0 | — |
case-03 | fail→fail | 21,060 | 17,662 | -16% | 1 | 1 | 0% | 2,064 | 3,747 | +82% | 0 | 0 | — |
case-06 | pass→pass | 21,288 | 28,291 | +33% | 1 | 1 | 0% | 3,889 | 5,198 | +34% | 0 | 0 | — |
case-07 | fail→pass | 23,056 | 7,464 | -68% | 1 | 1 | 0% | 2,081 | 1,782 | -14% | 0 | 0 | — |
case-08 | fail→pass | 9,564 | 6,465 | -32% | 1 | 1 | 0% | 1,604 | 891 | -44% | 0 | 0 | — |
case-09 | pass→fail | 23,014 | 22,260 | -3% | 1 | 1 | 0% | 1,994 | 2,740 | +37% | 0 | 0 | — |
case-10 | pass→pass | 10,769 | 7,539 | -30% | 1 | 1 | 0% | 1,776 | 1,860 | +5% | 0 | 0 | — |
case-11 | pass→pass | 11,296 | 11,453 | +1% | 1 | 1 | 0% | 1,305 | 1,787 | +37% | 0 | 0 | — |
case-12 | pass→pass | 8,396 | 9,907 | +18% | 1 | 1 | 0% | 1,274 | 1,896 | +49% | 0 | 0 | — |
case-13 | fail→fail | 8,359 | 6,780 | -19% | 1 | 1 | 0% | 1,370 | 1,870 | +36% | 0 | 0 | — |
case-14 | pass→pass | 21,749 | 7,503 | -66% | 1 | 1 | 0% | 1,115 | 1,927 | +73% | 0 | 0 | — |
case-15 | fail→pass | 10,565 | 4,807 | -55% | 1 | 1 | 0% | 1,992 | 1,333 | -33% | 0 | 0 | — |
case-16 | fail→pass | 9,950 | 4,476 | -55% | 1 | 1 | 0% | 1,529 | 1,041 | -32% | 0 | 0 | — |
case-17 | fail→pass | 7,893 | 2,743 | -65% | 1 | 1 | 0% | 1,298 | 1,093 | -16% | 0 | 0 | — |
case-18 | fail→pass | 3,211 | 2,763 | -14% | 1 | 1 | 0% | 532 | 911 | +71% | 0 | 0 | — |
case-19 | fail→pass | 15,253 | 9,764 | -36% | 1 | 1 | 0% | 2,654 | 2,270 | -14% | 0 | 0 | — |
case-20 | pass→pass | 14,826 | 10,434 | -30% | 1 | 1 | 0% | 2,430 | 2,388 | -2% | 0 | 0 | — |
case-21 | pass→pass | 12,712 | 15,284 | +20% | 1 | 1 | 0% | 2,104 | 2,894 | +38% | 0 | 0 | — |
case-22 | pass→pass | 15,038 | 13,211 | -12% | 1 | 1 | 0% | 2,604 | 2,750 | +6% | 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 +27 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.