Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Draft and submit a vulnerability report to the bug bounty platform. Reads scope.yaml for platform/program, uses brain + findings for content. Always drafts first for review.
.claude/skills/h-mmer-submit/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | 507% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -34% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 15% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -59% | 0% |
| case-04 | ✓→✗ | ▼ Worse | -52% | 0% |
Prepare and submit a report for finding: $ARGUMENTS
Workflow:
rules/identities.md to learn which env vars hold the researcher handle, email alias, and API token for the platform identified in step 1. NEVER hardcode a username or email; always reference the env-var symbol. If a required var is unset, abort with error: <VAR> is not set; refusing to guess and surface it to the user.scope.yaml to determine the platform and program handle.draft_report to create a platform-formatted draft:[Vuln Type] in [Component] allows [Impact] via [Vector]submit_report to submit.uv run python3 ../../tools/brain.py record <target> confirmed <technique> "Submitted as report #<id> on <platform>"IMPORTANT: NEVER submit without showing the draft and getting explicit user confirmation.
Submission is a controlled release.
Before asking for approval, verify:
/validate PASS or explicit accepted equivalent exists/quality score is acceptable and blocking issues are fixed/dupcheck result is included or intentionally skipped with reasonShow the user the final title, severity, platform, target asset, evidence list, and any residual risk. If anything changed after draft generation, re-run quality before submission.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-06 | pass→pass | 7,374 | 23,469 | +218% | 1 | 1 | 0% | 1,249 | 4,968 | +298% | 0 | 0 | — |
case-01 | fail→fail | 33,864 | 11,221 | -67% | 1 | 1 | 0% | 1,879 | 1,246 | -34% | 0 | 0 | — |
case-02 | fail→fail | 8,102 | 5,140 | -37% | 1 | 1 | 0% | 1,354 | 705 | -48% | 0 | 0 | — |
case-03 | fail→fail | 6,691 | 8,170 | +22% | 1 | 1 | 0% | 1,118 | 797 | -29% | 0 | 0 | — |
case-04 | pass→fail | 12,624 | 8,851 | -30% | 1 | 1 | 0% | 1,883 | 904 | -52% | 0 | 0 | — |
case-05 | fail→pass | 6,752 | 32,219 | +377% | 1 | 1 | 0% | 1,091 | 6,622 | +507% | 0 | 0 | — |
case-07 | pass→pass | 8,303 | 5,711 | -31% | 1 | 1 | 0% | 1,514 | 1,006 | -34% | 0 | 0 | — |
case-08 | pass→pass | 6,736 | 2,594 | -61% | 1 | 1 | 0% | 1,250 | 895 | -28% | 0 | 0 | — |
case-09 | pass→pass | 6,164 | 2,092 | -66% | 1 | 1 | 0% | 1,038 | 819 | -21% | 0 | 0 | — |
case-10 | fail→pass | 7,595 | 2,315 | -70% | 1 | 1 | 0% | 1,397 | 926 | -34% | 0 | 0 | — |
case-11 | pass→pass | 5,894 | 13,883 | +136% | 1 | 1 | 0% | 998 | 962 | -4% | 0 | 0 | — |
case-12 | fail→pass | 5,037 | 2,464 | -51% | 1 | 1 | 0% | 779 | 896 | +15% | 0 | 0 | — |
case-13 | pass→pass | 14,361 | 7,750 | -46% | 1 | 1 | 0% | 1,458 | 1,223 | -16% | 0 | 0 | — |
case-14 | fail→pass | 12,892 | 2,429 | -81% | 1 | 1 | 0% | 2,394 | 977 | -59% | 0 | 0 | — |
case-15 | pass→fail | 6,780 | 1,484 | -78% | 1 | 1 | 0% | 1,149 | 700 | -39% | 0 | 0 | — |
case-16 | pass→pass | 69,587 | 2,475 | -96% | 1 | 1 | 0% | 1,371 | 896 | -35% | 0 | 0 | — |
case-17 | pass→pass | 8,449 | 3,134 | -63% | 1 | 1 | 0% | 1,469 | 1,086 | -26% | 0 | 0 | — |
case-18 | pass→pass | 8,854 | 17,272 | +95% | 1 | 1 | 0% | 1,405 | 2,832 | +102% | 0 | 0 | — |
case-19 | pass→pass | 6,869 | 14,710 | +114% | 1 | 1 | 0% | 1,083 | 916 | -15% | 0 | 0 | — |
case-20 | pass→pass | 7,325 | 9,660 | +32% | 1 | 1 | 0% | 1,347 | 742 | -45% | 0 | 0 | — |
case-21 | fail→fail | 10,805 | 2,436 | -77% | 1 | 1 | 0% | 1,807 | 957 | -47% | 0 | 0 | — |
case-22 | pass→fail | 13,332 | 4,224 | -68% | 1 | 1 | 0% | 2,032 | 784 | -61% | 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 18 counted toward the lift figure. The other 4 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 +5 percentage points is the difference between those two pass rates over the 18 comparable cases. 5 cases got worse with the skill loaded, and they are 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.