Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Build deep exploit chains — dispatches chain-builder agent. Given bug A, recursively walks the chain graph. Usage: /chain (then describe bug A)
| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | — | — |
| case-16 | ✗→✗ | ＝ Same ✗ | — | — |
| case-18 | ✗→✗ | ＝ Same ✗ | — | — |
| case-20 | ✗→✗ | ＝ Same ✗ | — | — |
| case-15 | ✗→✗ | ＝ Same ✗ | — | — |
Build exploit chain from: $ARGUMENTS
uv run python3 $CLAUDE_PROJECT_DIR/tools/brain.py brief <target>
$ARGUMENTS contains a bug description → use itrules/chain-table.md — the capability→next-bug tablepolicy.md — extract policy preamble for the agentchain-builder agent (model: inherit) with:rules/chain-table.mdsearch_writeups "chain <bug class> escalation" if MCP availableuv run python3 $CLAUDE_PROJECT_DIR/tools/brain.py record <target> confirmed "chain: <summary>" "<full chain>"/validate then /reportuv run python3 $CLAUDE_PROJECT_DIR/tools/brain.py record <target> exhausted "chain from <bug A>" "<candidates tried>"No inline chain logic. No capability table. The chain-builder agent does all the work.
A chain is valuable only when each link grants a concrete capability.
Before dispatching, classify bug A as one capability:
Ask the chain-builder for three paths: fastest proof, highest impact, and safest policy-compliant path. Kill chains that require guessing, prohibited data access, or unbounded scanning. A reportable chain must include end-to-end reproduction, where link 2 consumes the capability from link 1 rather than merely coexisting with it.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-16 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-18 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-20 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-15 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-13 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-08 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-04 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-17 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-14 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-19 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-03 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-22 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-01 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-02 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-07 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-11 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-06 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-10 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-05 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-09 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-12 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-21 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
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 8 counted toward the lift figure. The other 14 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 8 comparable cases. 4 cases got worse with the skill loaded, and they are included in that figure.
The per-case answers from this run were removed by the retention sweep, so the case table below shows the verdicts without the text either arm produced. The counts above were recorded at the time and are unaffected. Answers are now kept for 180 days.
Other measured skills in the registry, with their headline benchmark lift.