Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when the user types /agentlas-search <query>, /agentlas search, or /hep-search to search across Local, Cloud, and Hub agent repositories.
.claude/skills/agentlas-ai-agentlas-search/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-23 | ✓→✓ | ＝ Same ✓ | 36% | 0% |
| case-24 | ✓→✓ | ＝ Same ✓ | 26% | 0% |
| case-25 | ✓→✓ | ＝ Same ✓ | 19% | 0% |
| case-26 | ✓→✓ | ＝ Same ✓ | 52% | 0% |
| case-01 | ✗→✗ | ＝ Same ✗ | -42% | 0% |
Identical to /skill:hep-search and /agentlas search <request>. Locate the sibling skill directory hep-search (i.e. ../hep-search/SKILL.md relative to this file, under the same kimi/skills/ root this skill was loaded from), read its SKILL.md, and follow its instructions exactly — treating everything typed after /skill:agentlas-search as that command's request.
Do not improvise a separate workflow and do not summarize hep-search/SKILL.md from memory; that file is the sole authority for this command's behavior.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 21,262 | 18,971 | -11% | 1 | 1 | 0% | 3,018 | 1,743 | -42% | 0 | 0 | — |
case-02 | fail→fail | 9,288 | 16,802 | +81% | 1 | 1 | 0% | 350 | 531 | +52% | 0 | 0 | — |
case-03 | fail→fail | 16,047 | 29,685 | +85% | 1 | 1 | 0% | 2,221 | 3,511 | +58% | 0 | 0 | — |
case-04 | fail→fail | 20,159 | 25,677 | +27% | 1 | 1 | 0% | 3,548 | 4,570 | +29% | 0 | 0 | — |
case-05 | fail→fail | 22,224 | 35,413 | +59% | 1 | 1 | 0% | 3,815 | 6,995 | +83% | 0 | 0 | — |
case-06 | fail→fail | 26,979 | 45,379 | +68% | 1 | 1 | 0% | 5,058 | 7,362 | +46% | 0 | 0 | — |
case-07 | fail→fail | 30,574 | 44,481 | +45% | 1 | 1 | 0% | 4,889 | 7,709 | +58% | 0 | 0 | — |
case-08 | fail→fail | 29,041 | 8,529 | -71% | 1 | 1 | 0% | 3,979 | 514 | -87% | 0 | 0 | — |
case-09 | fail→fail | 24,543 | 11,649 | -53% | 1 | 1 | 0% | 3,491 | 590 | -83% | 0 | 0 | — |
case-10 | fail→fail | 25,191 | 42,375 | +68% | 1 | 1 | 0% | 4,022 | 7,107 | +77% | 0 | 0 | — |
case-11 | fail→fail | 10,580 | 47,163 | +346% | 1 | 1 | 0% | 1,680 | 5,292 | +215% | 0 | 0 | — |
case-12 | fail→fail | 24,140 | 12,228 | -49% | 1 | 1 | 0% | 4,337 | 621 | -86% | 0 | 0 | — |
case-13 | fail→fail | 11,685 | 10,700 | -8% | 1 | 1 | 0% | 1,546 | 1,839 | +19% | 0 | 0 | — |
case-14 | fail→fail | 37,509 | 88,239 | +135% | 1 | 1 | 0% | 5,420 | 6,272 | +16% | 0 | 0 | — |
case-15 | fail→fail | 24,497 | 24,603 | +0% | 1 | 1 | 0% | 3,634 | 4,112 | +13% | 0 | 0 | — |
case-16 | fail→fail | 37,571 | 7,009 | -81% | 1 | 1 | 0% | 6,461 | 407 | -94% | 0 | 0 | — |
case-17 | fail→fail | 21,406 | 46,894 | +119% | 1 | 1 | 0% | 3,818 | 8,478 | +122% | 0 | 0 | — |
case-18 | fail→fail | 45,282 | 51,206 | +13% | 1 | 1 | 0% | 4,077 | 7,874 | +93% | 0 | 0 | — |
case-19 | fail→fail | 23,777 | 26,504 | +11% | 1 | 1 | 0% | 4,018 | 4,413 | +10% | 0 | 0 | — |
case-20 | fail→fail | 17,300 | 35,043 | +103% | 1 | 1 | 0% | 3,101 | 6,831 | +120% | 0 | 0 | — |
case-21 | fail→fail | 32,184 | 38,996 | +21% | 1 | 1 | 0% | 4,853 | 5,939 | +22% | 0 | 0 | — |
case-22 | fail→fail | 27,590 | 20,757 | -25% | 1 | 1 | 0% | 4,433 | 828 | -81% | 0 | 0 | — |
case-23 | pass→pass | 17,023 | 22,170 | +30% | 1 | 1 | 0% | 3,347 | 4,549 | +36% | 0 | 0 | — |
case-24 | pass→pass | 9,806 | 23,065 | +135% | 1 | 1 | 0% | 1,594 | 2,001 | +26% | 0 | 0 | — |
case-25 | pass→pass | 24,793 | 77,170 | +211% | 1 | 1 | 0% | 4,011 | 4,773 | +19% | 0 | 0 | — |
case-26 | pass→pass | 22,719 | 31,586 | +39% | 1 | 1 | 0% | 3,707 | 5,642 | +52% | 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. 26 cases were attempted, and 19 counted toward the lift figure. The other 7 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 0 percentage points is the difference between those two pass rates over the 19 comparable cases.
Other measured skills in the registry, with their headline benchmark lift.