Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Drafts and optionally sends WhatsApp replies with rater and HITL gates. Use when user says auto reply WhatsApp or enable WhatsApp assistant. Per-contact toggle; ethical guardrails and rate limits enforced. Triggers: auto reply W hatsApp, enable WhatsApp assistant, reply on WhatsApp.
.claude/skills/stijnman-whatsapp-auto-responder/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 115% | 0% |
| case-01 | ✗→✓ | ▲ Improved | 9% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -23% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -42% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -61% | 0% |
whatsapp-message-raterhitl-approverprivacy-redactorcron-scheduler| Failure | Response | |---------|----------| | Rate limit hit | Queue for user review; do not send. | | Bridge unavailable | Draft only; tell user to send manually. |
Input: User request matching triggers above. Output: Structured result per workflow with integrations invoked as needed.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-02 | fail→pass | 6,175 | 11,002 | +78% | 1 | 1 | 0% | 912 | 1,958 | +115% | 0 | 0 | — |
case-01 | fail→pass | 9,390 | 10,440 | +11% | 1 | 1 | 0% | 1,539 | 1,677 | +9% | 0 | 0 | — |
case-03 | fail→pass | 15,328 | 8,040 | -48% | 1 | 1 | 0% | 2,395 | 1,842 | -23% | 0 | 0 | — |
case-04 | pass→pass | 8,599 | 3,168 | -63% | 1 | 1 | 0% | 1,408 | 847 | -40% | 0 | 0 | — |
case-05 | fail→pass | 8,090 | 3,325 | -59% | 1 | 1 | 0% | 1,479 | 860 | -42% | 0 | 0 | — |
case-06 | pass→pass | 8,368 | 2,929 | -65% | 1 | 1 | 0% | 1,030 | 749 | -27% | 0 | 0 | — |
case-07 | pass→pass | 6,691 | 2,274 | -66% | 1 | 1 | 0% | 1,125 | 761 | -32% | 0 | 0 | — |
case-08 | pass→pass | 6,753 | 2,600 | -61% | 1 | 1 | 0% | 1,149 | 858 | -25% | 0 | 0 | — |
case-09 | fail→pass | 11,717 | 1,826 | -84% | 1 | 1 | 0% | 1,828 | 712 | -61% | 0 | 0 | — |
case-10 | pass→pass | 8,399 | 3,252 | -61% | 1 | 1 | 0% | 1,421 | 994 | -30% | 0 | 0 | — |
case-11 | pass→pass | 12,071 | 3,695 | -69% | 1 | 1 | 0% | 1,862 | 964 | -48% | 0 | 0 | — |
case-12 | pass→pass | 19,750 | 4,465 | -77% | 1 | 1 | 0% | 1,673 | 1,165 | -30% | 0 | 0 | — |
case-13 | fail→pass | 18,196 | 2,671 | -85% | 1 | 1 | 0% | 2,336 | 761 | -67% | 0 | 0 | — |
case-14 | pass→pass | 11,647 | 5,688 | -51% | 1 | 1 | 0% | 1,842 | 1,043 | -43% | 0 | 0 | — |
case-15 | pass→pass | 4,263 | 2,732 | -36% | 1 | 1 | 0% | 573 | 871 | +52% | 0 | 0 | — |
case-16 | pass→pass | 7,831 | 2,250 | -71% | 1 | 1 | 0% | 1,298 | 700 | -46% | 0 | 0 | — |
case-17 | pass→pass | 15,510 | 1,777 | -89% | 1 | 1 | 0% | 1,950 | 626 | -68% | 0 | 0 | — |
case-18 | pass→pass | 6,174 | 3,013 | -51% | 1 | 1 | 0% | 725 | 945 | +30% | 0 | 0 | — |
case-19 | pass→pass | 9,126 | 1,811 | -80% | 1 | 1 | 0% | 1,462 | 660 | -55% | 0 | 0 | — |
case-20 | fail→pass | 6,593 | 7,136 | +8% | 1 | 1 | 0% | 1,098 | 1,636 | +49% | 0 | 0 | — |
case-21 | fail→fail | 16,662 | 14,795 | -11% | 1 | 1 | 0% | 2,408 | 2,801 | +16% | 0 | 0 | — |
case-22 | fail→fail | 4,249 | 4,757 | +12% | 1 | 1 | 0% | 611 | 1,107 | +81% | 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 +32 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.