Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Prepare a structured discovery call plan for any prospect. Use when asked to prepare for a sales call, discovery call, prospect meeting, or first call with a potential customer. Produces a call brief with research, hypotheses, questions, and success criteria.
.claude/skills/mohitagw15856-discovery-call-prep/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 24% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 20% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 24% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 25% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 18% | 0% |
Produces a complete discovery call brief — research summary, call hypothesis, structured questions, and success criteria — so every call starts with context and ends with a clear next step.
Prospect: Company] | Contact: Name, Title] | Duration: X min]
Before the call write your best guess:
Write it down — then test it on the call.
"Here is what I was thinking for our X] minutes:
Does that work? Anything specific you would like to cover?"
Open with context (not a pitch):
Go deeper on pain:
Understand buying context:
Qualify on budget:
Close discovery:
This call is successful if we leave with:
This call is NOT successful if we only pitched and got "sounds interesting, send me some info."
"Based on what we discussed, the logical next step would be specific]. Does day/time] work?"
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-04 | pass→pass | 4,736 | 6,574 | +39% | 1 | 1 | 0% | 830 | 2,186 | +163% | 0 | 0 | — |
case-01 | fail→pass | 27,044 | 12,898 | -52% | 1 | 1 | 0% | 2,583 | 3,208 | +24% | 0 | 0 | — |
case-02 | fail→fail | 19,662 | 13,258 | -33% | 1 | 1 | 0% | 2,988 | 3,154 | +6% | 0 | 0 | — |
case-03 | fail→pass | 17,480 | 14,215 | -19% | 1 | 1 | 0% | 2,807 | 3,364 | +20% | 0 | 0 | — |
case-05 | pass→pass | 25,790 | 13,194 | -49% | 1 | 1 | 0% | 3,033 | 3,048 | +0% | 0 | 0 | — |
case-06 | pass→pass | 13,543 | 15,913 | +17% | 1 | 1 | 0% | 2,306 | 3,606 | +56% | 0 | 0 | — |
case-07 | fail→pass | 11,682 | 9,651 | -17% | 1 | 1 | 0% | 1,947 | 2,407 | +24% | 0 | 0 | — |
case-08 | pass→pass | 10,715 | 12,666 | +18% | 1 | 1 | 0% | 1,803 | 3,247 | +80% | 0 | 0 | — |
case-09 | fail→pass | 14,444 | 11,656 | -19% | 1 | 1 | 0% | 2,267 | 2,839 | +25% | 0 | 0 | — |
case-10 | pass→pass | 11,391 | 10,822 | -5% | 1 | 1 | 0% | 2,003 | 2,781 | +39% | 0 | 0 | — |
case-11 | pass→pass | 16,544 | 14,038 | -15% | 1 | 1 | 0% | 2,436 | 3,185 | +31% | 0 | 0 | — |
case-12 | pass→pass | 11,710 | 11,190 | -4% | 1 | 1 | 0% | 1,884 | 2,673 | +42% | 0 | 0 | — |
case-13 | fail→pass | 15,753 | 11,092 | -30% | 1 | 1 | 0% | 2,500 | 2,945 | +18% | 0 | 0 | — |
case-14 | fail→fail | 11,201 | 10,865 | -3% | 1 | 1 | 0% | 1,873 | 2,633 | +41% | 0 | 0 | — |
case-15 | fail→pass | 7,632 | 5,619 | -26% | 1 | 1 | 0% | 1,258 | 1,875 | +49% | 0 | 0 | — |
case-16 | fail→fail | 30,361 | 11,799 | -61% | 1 | 1 | 0% | 2,550 | 2,863 | +12% | 0 | 0 | — |
case-17 | fail→pass | 17,235 | 12,194 | -29% | 1 | 1 | 0% | 2,575 | 2,754 | +7% | 0 | 0 | — |
case-18 | fail→pass | 15,380 | 11,798 | -23% | 1 | 1 | 0% | 2,256 | 2,694 | +19% | 0 | 0 | — |
case-19 | pass→pass | 12,837 | 9,334 | -27% | 1 | 1 | 0% | 1,950 | 2,343 | +20% | 0 | 0 | — |
case-20 | fail→pass | 5,716 | 2,345 | -59% | 1 | 1 | 0% | 930 | 1,331 | +43% | 0 | 0 | — |
case-21 | pass→pass | 16,933 | 9,947 | -41% | 1 | 1 | 0% | 2,476 | 2,546 | +3% | 0 | 0 | — |
case-22 | fail→pass | 24,731 | 14,624 | -41% | 1 | 1 | 0% | 3,755 | 3,464 | -8% | 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 +45 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.