Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Ask a quick side question about your current work without derailing the main task. Answers from existing conversation context only — no tool calls, no file reads, single concise response. Use when you need a fast answer from what is already in this session.
.claude/skills/mkurman-btw/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 30% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -1% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 101% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 15% | 0% |
| case-08 | ✓→✗ | ▼ Worse | 110% | 0% |
<objective> Answer a quick side question using only what is already present in the current conversation context. Do not read files, run commands, search, or use any tools. Give a single, concise response and return focus to the main work. </objective>
<behavior> This is a side question, not a task.
</behavior>
<quick_start> Parse the argument after /btw as the question. Answer it directly from context.
If no argument is provided, ask: "What did you want to know?"
If the question cannot be answered from current context (requires reading a file, running a command, or information not yet in the session), respond with: "I'd need to read X / run Y / look up Z] to answer that — ask it as a normal prompt when you're ready." </quick_start>
<examples> Good uses of /btw:
/btw what was the name of that config file again? → answers from files already read in session/btw which branch are we on? → answers from git context already established/btw did we already handle the null case in that function? → answers from code already reviewed/btw what model does this use? → answers from code or config already in contextNot a good fit for /btw (suggest normal prompt):
</examples>
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 3,301 | 2,516 | -24% | 1 | 1 | 0% | 394 | 795 | +102% | 0 | 0 | — |
case-02 | pass→pass | 2,177 | 1,723 | -21% | 1 | 1 | 0% | 296 | 670 | +126% | 0 | 0 | — |
case-03 | fail→pass | 3,426 | 1,074 | -69% | 1 | 1 | 0% | 390 | 507 | +30% | 0 | 0 | — |
case-04 | fail→fail | 3,301 | 1,810 | -45% | 1 | 1 | 0% | 501 | 675 | +35% | 0 | 0 | — |
case-18 | fail→fail | 7,202 | 2,177 | -70% | 1 | 1 | 0% | 446 | 731 | +64% | 0 | 0 | — |
case-05 | fail→fail | 2,639 | 2,140 | -19% | 1 | 1 | 0% | 329 | 710 | +116% | 0 | 0 | — |
case-06 | fail→fail | 5,138 | 2,505 | -51% | 1 | 1 | 0% | 750 | 823 | +10% | 0 | 0 | — |
case-07 | fail→fail | 4,040 | 2,049 | -49% | 1 | 1 | 0% | 338 | 723 | +114% | 0 | 0 | — |
case-08 | pass→fail | 3,264 | 2,308 | -29% | 1 | 1 | 0% | 370 | 778 | +110% | 0 | 0 | — |
case-09 | fail→pass | 4,371 | 1,897 | -57% | 1 | 1 | 0% | 689 | 685 | -1% | 0 | 0 | — |
case-10 | fail→pass | 2,501 | 1,798 | -28% | 1 | 1 | 0% | 321 | 645 | +101% | 0 | 0 | — |
case-11 | fail→pass | 3,872 | 1,635 | -58% | 1 | 1 | 0% | 574 | 658 | +15% | 0 | 0 | — |
case-12 | fail→fail | 4,716 | 2,225 | -53% | 1 | 1 | 0% | 623 | 735 | +18% | 0 | 0 | — |
case-13 | pass→pass | 2,698 | 1,724 | -36% | 1 | 1 | 0% | 351 | 685 | +95% | 0 | 0 | — |
case-14 | fail→fail | 3,010 | 1,897 | -37% | 1 | 1 | 0% | 336 | 701 | +109% | 0 | 0 | — |
case-15 | fail→fail | 3,315 | 2,660 | -20% | 1 | 1 | 0% | 458 | 834 | +82% | 0 | 0 | — |
case-16 | fail→fail | 4,359 | 1,986 | -54% | 1 | 1 | 0% | 630 | 722 | +15% | 0 | 0 | — |
case-17 | fail→fail | 2,770 | 2,294 | -17% | 1 | 1 | 0% | 334 | 717 | +115% | 0 | 0 | — |
case-19 | fail→fail | 3,508 | 2,027 | -42% | 1 | 1 | 0% | 487 | 675 | +39% | 0 | 0 | — |
case-20 | pass→fail | 5,670 | 9,583 | +69% | 1 | 1 | 0% | 956 | 1,601 | +67% | 0 | 0 | — |
case-21 | fail→fail | 7,851 | 7,161 | -9% | 1 | 1 | 0% | 563 | 1,259 | +124% | 0 | 0 | — |
case-22 | pass→fail | 11,339 | 14,303 | +26% | 1 | 1 | 0% | 2,127 | 2,352 | +11% | 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. The headline lift of 0 percentage points is the difference between those two pass rates over the 22 comparable cases. 4 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.