Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Reach back out to a friend or person you've lost touch with — past the awkwardness of the gap — with a message that reopens the door warmly. Use when asked how do I reconnect with an old friend, it's been too long and it's awkward, reach out to someone I drifted from, or message someone I lost touch with. Produces a read on why the awkwardness is smaller than it feels, a warm reach-out message that acknowledges the gap without over-apologizing, a specific hook (a memory, a reason, a simple 'you
.claude/skills/mohitagw15856-reconnect-with-someone/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 55% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 46% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 16% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 15% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 20% | 0% |
Most lost friendships didn't end — they just drifted, and then the gap itself became the barrier ("it's been so long, it'd be weird now"). But the other person almost always feels the same and is glad you reached out. This gets you past the awkwardness with a warm message that acknowledges the gap lightly, gives a genuine reason for reaching out, and reopens the door — because reconnecting usually just needs one person brave enough to text first.
Ask for these if not provided:
The reframe: why they'll likely be glad, not annoyed — the awkwardness is mostly yours]. Your message > Warm, light acknowledgment of the gap + a specific genuine hook + an easy opening — not a heavy apology or a wall of text].
The hook you're using: memory / reminder / update / "you crossed my mind"]. Then: once they reply, suggest a low-key call/meet-up]. Keep it: no guilt, warm, forward-looking.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 16,546 | 13,512 | -18% | 1 | 1 | 0% | 1,524 | 2,360 | +55% | 0 | 0 | — |
case-02 | fail→pass | 19,369 | 14,379 | -26% | 1 | 1 | 0% | 1,729 | 2,526 | +46% | 0 | 0 | — |
case-03 | fail→pass | 20,441 | 17,328 | -15% | 1 | 1 | 0% | 2,300 | 2,674 | +16% | 0 | 0 | — |
case-04 | fail→fail | 29,776 | 15,135 | -49% | 1 | 1 | 0% | 1,607 | 2,250 | +40% | 0 | 0 | — |
case-05 | fail→pass | 20,114 | 14,884 | -26% | 1 | 1 | 0% | 1,977 | 2,264 | +15% | 0 | 0 | — |
case-06 | fail→pass | 15,286 | 11,864 | -22% | 1 | 1 | 0% | 2,053 | 2,469 | +20% | 0 | 0 | — |
case-07 | fail→pass | 9,998 | 8,421 | -16% | 1 | 1 | 0% | 1,469 | 2,210 | +50% | 0 | 0 | — |
case-08 | fail→pass | 8,266 | 7,332 | -11% | 1 | 1 | 0% | 1,293 | 2,140 | +66% | 0 | 0 | — |
case-09 | fail→pass | 12,142 | 10,573 | -13% | 1 | 1 | 0% | 1,500 | 2,552 | +70% | 0 | 0 | — |
case-10 | fail→pass | 15,112 | 8,956 | -41% | 1 | 1 | 0% | 1,248 | 2,130 | +71% | 0 | 0 | — |
case-11 | pass→pass | 11,710 | 16,516 | +41% | 1 | 1 | 0% | 1,466 | 3,059 | +109% | 0 | 0 | — |
case-12 | pass→pass | 10,171 | 17,966 | +77% | 1 | 1 | 0% | 1,587 | 3,399 | +114% | 0 | 0 | — |
case-13 | pass→pass | 9,315 | 10,354 | +11% | 1 | 1 | 0% | 1,261 | 2,529 | +101% | 0 | 0 | — |
case-14 | fail→pass | 10,651 | 8,471 | -20% | 1 | 1 | 0% | 1,589 | 2,248 | +41% | 0 | 0 | — |
case-15 | fail→pass | 9,807 | 9,364 | -5% | 1 | 1 | 0% | 1,326 | 2,202 | +66% | 0 | 0 | — |
case-16 | fail→pass | 8,186 | 7,394 | -10% | 1 | 1 | 0% | 1,227 | 1,998 | +63% | 0 | 0 | — |
case-17 | fail→pass | 12,090 | 6,569 | -46% | 1 | 1 | 0% | 1,719 | 2,028 | +18% | 0 | 0 | — |
case-18 | fail→pass | 9,725 | 9,135 | -6% | 1 | 1 | 0% | 1,261 | 2,219 | +76% | 0 | 0 | — |
case-19 | fail→pass | 11,776 | 9,559 | -19% | 1 | 1 | 0% | 1,368 | 2,203 | +61% | 0 | 0 | — |
case-20 | fail→pass | 12,228 | 10,566 | -14% | 1 | 1 | 0% | 1,682 | 2,208 | +31% | 0 | 0 | — |
case-21 | fail→pass | 7,013 | 7,162 | +2% | 1 | 1 | 0% | 1,125 | 2,026 | +80% | 0 | 0 | — |
case-22 | fail→pass | 8,942 | 7,675 | -14% | 1 | 1 | 0% | 1,328 | 2,052 | +55% | 0 | 0 | — |
case-23 | fail→pass | 8,642 | 14,084 | +63% | 1 | 1 | 0% | 1,294 | 2,094 | +62% | 0 | 0 | — |
case-24 | fail→fail | 7,584 | 9,255 | +22% | 1 | 1 | 0% | 1,110 | 2,079 | +87% | 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. 24 cases were attempted. The headline lift of +79 percentage points is the difference between those two pass rates over the 24 comparable cases. 1 case got worse with the skill loaded, and it is 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.