Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Read this skill when the user uses LastPass.
.claude/skills/kunanonj-aside-lastpass/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -51% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -60% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -68% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -46% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -59% | 0% |
LastPass autofill options may appear after focusing a login form field.
Then check whether the snapshot includes a LastPass autofill menu or iframe [origin="LastPass"].
Open the LastPass extension page only to unlock LastPass with Aside PWM:
chrome-extension://hdokiejnpimakedhajhdlcegeplioahd/popup.html
Then run:
jsawait passwordManager.unlockExternalPasswordManager(page, 'lastpass');
After unlocking LastPass:
Do not use the LastPass extension page as the primary place to search and autofill website logins. Use the original sign-in page's LastPass autofill menu for website login autofill.
If Aside PWM is locked, ask the user to unlock either Aside PWM or LastPass.
If no saved LastPass unlock item is available in Aside PWM, fall back to other password manager options. Ask the user to unlock LastPass directly only as the last resort.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 3,021 | 5,113 | +69% | 1 | 1 | 0% | 434 | 593 | +37% | 0 | 0 | — |
case-02 | fail→fail | 6,104 | 13,973 | +129% | 1 | 1 | 0% | 608 | 817 | +34% | 0 | 0 | — |
case-03 | fail→fail | 4,465 | 8,217 | +84% | 1 | 1 | 0% | 720 | 748 | +4% | 0 | 0 | — |
case-04 | fail→pass | 10,184 | 3,026 | -70% | 1 | 1 | 0% | 1,598 | 783 | -51% | 0 | 0 | — |
case-05 | fail→pass | 8,620 | 4,068 | -53% | 1 | 1 | 0% | 1,629 | 658 | -60% | 0 | 0 | — |
case-06 | pass→pass | 7,820 | 3,915 | -50% | 1 | 1 | 0% | 1,253 | 810 | -35% | 0 | 0 | — |
case-07 | pass→pass | 9,030 | 1,599 | -82% | 1 | 1 | 0% | 1,462 | 601 | -59% | 0 | 0 | — |
case-08 | fail→pass | 11,442 | 1,613 | -86% | 1 | 1 | 0% | 2,069 | 659 | -68% | 0 | 0 | — |
case-09 | pass→pass | 6,060 | 1,898 | -69% | 1 | 1 | 0% | 839 | 684 | -18% | 0 | 0 | — |
case-10 | pass→pass | 5,514 | 1,963 | -64% | 1 | 1 | 0% | 1,014 | 801 | -21% | 0 | 0 | — |
case-11 | pass→pass | 7,276 | 1,792 | -75% | 1 | 1 | 0% | 1,363 | 718 | -47% | 0 | 0 | — |
case-12 | pass→pass | 10,297 | 3,193 | -69% | 1 | 1 | 0% | 1,708 | 612 | -64% | 0 | 0 | — |
case-13 | fail→pass | 7,799 | 3,449 | -56% | 1 | 1 | 0% | 1,353 | 729 | -46% | 0 | 0 | — |
case-14 | fail→pass | 10,605 | 4,518 | -57% | 1 | 1 | 0% | 1,794 | 734 | -59% | 0 | 0 | — |
case-15 | fail→pass | 9,020 | 3,293 | -63% | 1 | 1 | 0% | 1,540 | 646 | -58% | 0 | 0 | — |
case-16 | fail→pass | 10,062 | 1,382 | -86% | 1 | 1 | 0% | 1,655 | 694 | -58% | 0 | 0 | — |
case-17 | pass→pass | 6,006 | 2,532 | -58% | 1 | 1 | 0% | 1,114 | 933 | -16% | 0 | 0 | — |
case-18 | pass→pass | 5,621 | 40,960 | +629% | 1 | 1 | 0% | 975 | 864 | -11% | 0 | 0 | — |
case-19 | pass→pass | 4,860 | 10,734 | +121% | 1 | 1 | 0% | 684 | 742 | +8% | 0 | 0 | — |
case-20 | pass→pass | 7,378 | 2,243 | -70% | 1 | 1 | 0% | 1,325 | 827 | -38% | 0 | 0 | — |
case-21 | pass→pass | 15,439 | 10,882 | -30% | 1 | 1 | 0% | 2,993 | 2,541 | -15% | 0 | 0 | — |
case-22 | fail→pass | 13,320 | 10,560 | -21% | 1 | 1 | 0% | 2,058 | 1,591 | -23% | 0 | 0 | — |
case-23 | pass→pass | 13,352 | 8,218 | -38% | 1 | 1 | 0% | 2,381 | 1,873 | -21% | 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. 23 cases were attempted, and 20 counted toward the lift figure. The other 3 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 +35 percentage points is the difference between those two pass rates over the 20 comparable cases.
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.