Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Run a quick security scan on a target. Consults the Brain first, validates scope, runs passive recon + vuln scan in parallel.
.claude/skills/h-mmer-quickscan/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 110% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -41% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -31% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -65% | 0% |
ALL agents dispatched by this command MUST use in the subagent dispatch tool call.
Run a quick security assessment on: $ARGUMENTS
Workflow:
uv run python3 ../../tools/brain.py brief $ARGUMENTS — check what we already know. Note exhausted areas.uv run python3 ../../tools/scope_check.py $ARGUMENTS — if out of scope, STOP.recon agent with passive-only depth, passing brain context about known subdomains/techconfig-auditor agent for headers, CSP, CORS, TLS, cookiesuv run python3 ../../tools/brain.py record <target> <status> <technique> <details>uv run python3 ../../tools/brain.py log "quickscan completed on $ARGUMENTS"Quickscan should answer "is there obvious money or obvious risk here in 30 minutes?"
Never report from quickscan alone unless the proof is already complete. Promote strong leads to /hunt, /validate, or /chain.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 8,293 | 6,062 | -27% | 1 | 1 | 0% | 879 | 1,023 | +16% | 0 | 0 | — |
case-02 | fail→fail | 7,558 | 7,415 | -2% | 1 | 1 | 0% | 580 | 1,002 | +73% | 0 | 0 | — |
case-03 | fail→fail | 7,522 | 9,042 | +20% | 1 | 1 | 0% | 735 | 1,664 | +126% | 0 | 0 | — |
case-04 | fail→fail | 5,546 | 4,634 | -16% | 1 | 1 | 0% | 521 | 813 | +56% | 0 | 0 | — |
case-05 | fail→fail | 12,336 | 6,386 | -48% | 1 | 1 | 0% | 1,900 | 770 | -59% | 0 | 0 | — |
case-06 | pass→pass | 7,426 | 7,356 | -1% | 1 | 1 | 0% | 842 | 1,235 | +47% | 0 | 0 | — |
case-07 | fail→pass | 9,494 | 8,545 | -10% | 1 | 1 | 0% | 1,054 | 1,190 | +13% | 0 | 0 | — |
case-08 | fail→pass | 7,178 | 9,178 | +28% | 1 | 1 | 0% | 663 | 1,392 | +110% | 0 | 0 | — |
case-09 | pass→pass | 12,974 | 13,855 | +7% | 1 | 1 | 0% | 1,625 | 2,028 | +25% | 0 | 0 | — |
case-10 | fail→pass | 14,015 | 5,210 | -63% | 1 | 1 | 0% | 2,385 | 1,408 | -41% | 0 | 0 | — |
case-11 | pass→pass | 12,471 | 13,369 | +7% | 1 | 1 | 0% | 1,228 | 1,321 | +8% | 0 | 0 | — |
case-12 | fail→pass | 6,851 | 1,989 | -71% | 1 | 1 | 0% | 1,131 | 779 | -31% | 0 | 0 | — |
case-13 | fail→pass | 9,252 | 1,552 | -83% | 1 | 1 | 0% | 1,807 | 637 | -65% | 0 | 0 | — |
case-14 | pass→pass | 11,775 | 5,708 | -52% | 1 | 1 | 0% | 1,942 | 1,391 | -28% | 0 | 0 | — |
case-15 | fail→pass | 9,492 | 1,690 | -82% | 1 | 1 | 0% | 1,646 | 652 | -60% | 0 | 0 | — |
case-16 | fail→pass | 10,411 | 1,908 | -82% | 1 | 1 | 0% | 1,735 | 724 | -58% | 0 | 0 | — |
case-17 | pass→pass | 5,906 | 2,281 | -61% | 1 | 1 | 0% | 979 | 716 | -27% | 0 | 0 | — |
case-18 | fail→pass | 8,741 | 3,340 | -62% | 1 | 1 | 0% | 1,422 | 943 | -34% | 0 | 0 | — |
case-19 | pass→pass | 10,126 | 3,643 | -64% | 1 | 1 | 0% | 1,657 | 1,030 | -38% | 0 | 0 | — |
case-20 | fail→pass | 13,702 | 3,397 | -75% | 1 | 1 | 0% | 2,228 | 671 | -70% | 0 | 0 | — |
case-21 | fail→pass | 11,269 | 2,054 | -82% | 1 | 1 | 0% | 1,502 | 687 | -54% | 0 | 0 | — |
case-22 | fail→pass | 26,295 | 2,245 | -91% | 1 | 1 | 0% | 4,789 | 772 | -84% | 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, and 21 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 +50 percentage points is the difference between those two pass rates over the 21 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.