Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Draft clear, warm, professional messages to parents or guardians — progress notes, concerns, positive news, behaviour issues, or meeting requests. Use when asked to email a parent, write home about a student, raise a concern with a guardian, or share an update. Produces a ready-to-send message that is specific, partnership-oriented, and constructive — never accusatory — with the tone matched to the situation.
.claude/skills/mohitagw15856-parent-communication/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 104% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 37% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 91% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 3% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 64% | 0% |
Messages home set the tone for the whole relationship. The best ones are specific, lead with care for the child, frame issues as a shared problem to solve, and always include a next step. This skill writes them.
Given the situation, write the full message anyway using a placeholder-free template (e.g. "Alex" / "your child" rather than "student name]" only where the teacher must personalise — keep those to an obvious minimum and mark them clearly). Match the tone to the purpose.
Ask for (if not already provided):
A ready-to-send message:
For a sensitive issue, also give:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 7,994 | 9,148 | +14% | 1 | 1 | 0% | 1,396 | 2,008 | +44% | 0 | 0 | — |
case-02 | pass→pass | 7,864 | 8,635 | +10% | 1 | 1 | 0% | 1,393 | 1,968 | +41% | 0 | 0 | — |
case-03 | pass→pass | 9,632 | 9,141 | -5% | 1 | 1 | 0% | 1,715 | 1,990 | +16% | 0 | 0 | — |
case-04 | fail→pass | 3,601 | 4,159 | +15% | 1 | 1 | 0% | 628 | 1,279 | +104% | 0 | 0 | — |
case-05 | fail→pass | 7,465 | 7,849 | +5% | 1 | 1 | 0% | 1,353 | 1,850 | +37% | 0 | 0 | — |
case-06 | fail→pass | 6,767 | 9,833 | +45% | 1 | 1 | 0% | 1,185 | 2,262 | +91% | 0 | 0 | — |
case-07 | fail→fail | 7,707 | 7,997 | +4% | 1 | 1 | 0% | 1,412 | 1,917 | +36% | 0 | 0 | — |
case-08 | pass→pass | 8,811 | 7,965 | -10% | 1 | 1 | 0% | 1,532 | 1,708 | +11% | 0 | 0 | — |
case-09 | pass→pass | 7,091 | 6,359 | -10% | 1 | 1 | 0% | 1,144 | 1,529 | +34% | 0 | 0 | — |
case-10 | pass→pass | 10,331 | 9,696 | -6% | 1 | 1 | 0% | 1,613 | 1,912 | +19% | 0 | 0 | — |
case-11 | pass→pass | 5,848 | 7,783 | +33% | 1 | 1 | 0% | 971 | 1,674 | +72% | 0 | 0 | — |
case-12 | fail→pass | 9,850 | 8,454 | -14% | 1 | 1 | 0% | 1,776 | 1,838 | +3% | 0 | 0 | — |
case-13 | pass→pass | 7,078 | 8,497 | +20% | 1 | 1 | 0% | 1,232 | 1,924 | +56% | 0 | 0 | — |
case-14 | pass→pass | 5,225 | 3,201 | -39% | 1 | 1 | 0% | 942 | 981 | +4% | 0 | 0 | — |
case-19 | pass→pass | 13,809 | 10,979 | -20% | 1 | 1 | 0% | 2,182 | 2,415 | +11% | 0 | 0 | — |
case-15 | pass→pass | 6,861 | 7,119 | +4% | 1 | 1 | 0% | 1,197 | 1,703 | +42% | 0 | 0 | — |
case-16 | fail→pass | 7,730 | 10,071 | +30% | 1 | 1 | 0% | 1,340 | 2,202 | +64% | 0 | 0 | — |
case-17 | fail→pass | 4,943 | 7,706 | +56% | 1 | 1 | 0% | 907 | 1,759 | +94% | 0 | 0 | — |
case-18 | pass→pass | 6,491 | 8,657 | +33% | 1 | 1 | 0% | 1,129 | 1,977 | +75% | 0 | 0 | — |
case-20 | fail→fail | 8,219 | 11,992 | +46% | 1 | 1 | 0% | 1,403 | 2,434 | +73% | 0 | 0 | — |
case-21 | pass→pass | 16,638 | 15,918 | -4% | 1 | 1 | 0% | 2,737 | 3,140 | +15% | 0 | 0 | — |
case-22 | pass→pass | 11,272 | 8,130 | -28% | 1 | 1 | 0% | 1,795 | 1,766 | -2% | 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 +27 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.