Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Set a boundary with in-laws or extended family kindly and clearly — the words to say, a united-front approach with your partner, and a plan for the pushback. Use when asked how to set a boundary with my in-laws, my mother-in-law keeps [X], deal with overbearing family, or what to say to my partner's family. Produces a read on the actual issue, a warm-but-firm script for the specific situation, a partner-alignment plan (the couple presents together), how to hold the line when they push back, and
.claude/skills/mohitagw15856-in-law-boundary-scripts/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-16 | ✗→✓ | ▲ Improved | 21% | 0% |
| case-01 | ✓→✓ | = Same ✓ | 2% | 0% |
| case-02 | ✓→✓ | = Same ✓ | 20% | 0% |
| case-03 | ✓→✓ | = Same ✓ | 21% | 0% |
| case-04 | ✓→✓ | = Same ✓ | 60% | 0% |
Boundaries with extended family are hard because you want to protect your peace and the relationships. The trick is to be warm and clear at once, and — crucially — for the couple to be united, with each partner leading with their own family. This gives you the actual words for the situation, a plan to align with your partner first, and a way to hold the line when the guilt-trips come.
Ask for these if not provided:
What you need: the specific behavior change]. Align first (with your partner): agree the boundary · who delivers it — ideally their own parent].
Script > Genuine appreciation] + the boundary, clearly] + the concrete ask]. — warm, firm, no over-apology.
When they push back: "calm restatement]" — hold, don't debate. Keep it: about the behavior and going forward, not their character.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 24,703 | 20,467 | -17% | 1 | 1 | 0% | 2,974 | 3,037 | +2% | 0 | 0 | — |
case-02 | pass→pass | 24,108 | 20,805 | -14% | 1 | 1 | 0% | 2,579 | 3,099 | +20% | 0 | 0 | — |
case-03 | pass→pass | 21,779 | 27,463 | +26% | 1 | 1 | 0% | 2,539 | 3,067 | +21% | 0 | 0 | — |
case-04 | pass→pass | 21,308 | 21,814 | +2% | 1 | 1 | 0% | 1,892 | 3,030 | +60% | 0 | 0 | — |
case-05 | pass→pass | 18,763 | 15,553 | -17% | 1 | 1 | 0% | 1,812 | 2,999 | +66% | 0 | 0 | — |
case-06 | pass→pass | 23,094 | 16,516 | -28% | 1 | 1 | 0% | 1,976 | 2,576 | +30% | 0 | 0 | — |
case-07 | pass→pass | 16,697 | 18,112 | +8% | 1 | 1 | 0% | 2,391 | 2,965 | +24% | 0 | 0 | — |
case-08 | pass→pass | 16,972 | 15,139 | -11% | 1 | 1 | 0% | 1,790 | 2,506 | +40% | 0 | 0 | — |
case-09 | pass→pass | 18,424 | 30,933 | +68% | 1 | 1 | 0% | 2,023 | 2,722 | +35% | 0 | 0 | — |
case-10 | pass→pass | 20,612 | 16,992 | -18% | 1 | 1 | 0% | 1,946 | 2,770 | +42% | 0 | 0 | — |
case-11 | fail→fail | 18,201 | 14,830 | -19% | 1 | 1 | 0% | 1,931 | 2,372 | +23% | 0 | 0 | — |
case-12 | pass→pass | 21,299 | 15,817 | -26% | 1 | 1 | 0% | 2,211 | 2,596 | +17% | 0 | 0 | — |
case-13 | pass→pass | 22,083 | 22,090 | +0% | 1 | 1 | 0% | 2,150 | 3,237 | +51% | 0 | 0 | — |
case-14 | pass→pass | 21,158 | 21,152 | -0% | 1 | 1 | 0% | 2,232 | 2,865 | +28% | 0 | 0 | — |
case-15 | pass→pass | 29,818 | 19,972 | -33% | 1 | 1 | 0% | 2,326 | 2,774 | +19% | 0 | 0 | — |
case-16 | fail→pass | 24,203 | 16,292 | -33% | 1 | 1 | 0% | 2,555 | 3,100 | +21% | 0 | 0 | — |
case-17 | pass→pass | 24,525 | 18,870 | -23% | 1 | 1 | 0% | 2,233 | 2,892 | +30% | 0 | 0 | — |
case-18 | pass→pass | 23,633 | 18,129 | -23% | 1 | 1 | 0% | 1,319 | 1,914 | +45% | 0 | 0 | — |
case-19 | pass→pass | 14,220 | 16,665 | +17% | 1 | 1 | 0% | 2,119 | 2,775 | +31% | 0 | 0 | — |
case-20 | pass→pass | 22,216 | 16,233 | -27% | 1 | 1 | 0% | 2,755 | 3,399 | +23% | 0 | 0 | — |
case-21 | pass→pass | 24,593 | 24,159 | -2% | 1 | 1 | 0% | 2,457 | 3,588 | +46% | 0 | 0 | — |
case-22 | pass→pass | 15,224 | 18,916 | +24% | 1 | 1 | 0% | 2,252 | 3,003 | +33% | 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 +5 percentage points is the difference between those two pass rates over the 22 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.