Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Run a calm, regular relationship check-in with your partner — a structured 'how are we doing' conversation that catches small things before they become big ones. Use when asked how to check in with my partner, we need to talk about our relationship, set up a relationship check-in, or improve communication with my partner. Produces a simple check-in structure (appreciations, what's working, what needs attention, needs and asks), ground rules that keep it safe not combative, prompts to surface the
.claude/skills/mohitagw15856-relationship-check-in/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-09 | ✗→✓ | ▲ Improved | 35% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 40% | 0% |
| case-22 | ✗→✓ | ▲ Improved | 59% | 0% |
| case-04 | ✓→✗ | ▼ Worse | 100% | 0% |
| case-05 | ✓→✗ | ▼ Worse | 90% | 0% |
Most relationship problems aren't sudden — they're small unspoken things that pile up. A regular, low-drama check-in surfaces them early, in a format that feels like teamwork rather than a confrontation. This gives you a simple structure, ground rules that keep it safe, and prompts that get past "we're fine" to what's actually going on.
Ask for these if not provided:
Ground rules: no blame/scorekeeping · listen to understand · one topic · pause if heated.
The structure
Prompts to go deeper: connection · stress · intimacy · fairness/chores · future]. Cadence: regular + short] in a low-key setting]. Bigger than a check-in? recurring issue / trust / serious conflict → consider a counselor].
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 19,756 | 18,291 | -7% | 1 | 1 | 0% | 2,506 | 3,451 | +38% | 0 | 0 | — |
case-02 | pass→pass | 24,533 | 18,521 | -25% | 1 | 1 | 0% | 3,188 | 3,079 | -3% | 0 | 0 | — |
case-03 | pass→pass | 26,176 | 21,015 | -20% | 1 | 1 | 0% | 3,008 | 3,325 | +11% | 0 | 0 | — |
case-04 | pass→fail | 26,672 | 38,483 | +44% | 1 | 1 | 0% | 3,431 | 6,871 | +100% | 0 | 0 | — |
case-05 | pass→fail | 21,769 | 30,322 | +39% | 1 | 1 | 0% | 2,929 | 5,559 | +90% | 0 | 0 | — |
case-06 | pass→pass | 16,218 | 11,472 | -29% | 1 | 1 | 0% | 1,721 | 2,902 | +69% | 0 | 0 | — |
case-07 | pass→fail | 20,054 | 36,247 | +81% | 1 | 1 | 0% | 2,643 | 6,082 | +130% | 0 | 0 | — |
case-08 | pass→pass | 20,009 | 17,918 | -10% | 1 | 1 | 0% | 2,260 | 3,024 | +34% | 0 | 0 | — |
case-09 | fail→pass | 20,339 | 18,275 | -10% | 1 | 1 | 0% | 2,271 | 3,075 | +35% | 0 | 0 | — |
case-10 | fail→pass | 22,118 | 19,065 | -14% | 1 | 1 | 0% | 2,232 | 3,129 | +40% | 0 | 0 | — |
case-11 | pass→pass | 19,034 | 17,735 | -7% | 1 | 1 | 0% | 2,201 | 2,690 | +22% | 0 | 0 | — |
case-12 | pass→pass | 15,249 | 20,429 | +34% | 1 | 1 | 0% | 2,147 | 3,133 | +46% | 0 | 0 | — |
case-13 | pass→pass | 20,535 | 11,253 | -45% | 1 | 1 | 0% | 2,001 | 2,844 | +42% | 0 | 0 | — |
case-14 | pass→pass | 14,060 | 14,191 | +1% | 1 | 1 | 0% | 1,852 | 2,843 | +54% | 0 | 0 | — |
case-15 | pass→pass | 13,695 | 11,339 | -17% | 1 | 1 | 0% | 2,076 | 2,810 | +35% | 0 | 0 | — |
case-20 | pass→pass | 17,461 | 13,807 | -21% | 1 | 1 | 0% | 1,869 | 2,821 | +51% | 0 | 0 | — |
case-16 | pass→pass | 14,002 | 15,565 | +11% | 1 | 1 | 0% | 1,144 | 2,430 | +112% | 0 | 0 | — |
case-17 | pass→pass | 17,117 | 18,855 | +10% | 1 | 1 | 0% | 1,680 | 2,755 | +64% | 0 | 0 | — |
case-18 | pass→pass | 16,356 | 17,640 | +8% | 1 | 1 | 0% | 1,612 | 2,677 | +66% | 0 | 0 | — |
case-19 | pass→pass | 15,608 | 17,811 | +14% | 1 | 1 | 0% | 2,195 | 3,026 | +38% | 0 | 0 | — |
case-21 | pass→pass | 21,227 | 14,204 | -33% | 1 | 1 | 0% | 2,455 | 3,073 | +25% | 0 | 0 | — |
case-22 | fail→pass | 16,169 | 8,190 | -49% | 1 | 1 | 0% | 1,545 | 2,450 | +59% | 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. 3 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.