Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Build the case to win someone over to a decision, idea, or change. Use when asked to persuade someone, build a case for an idea, get buy-in, win over a skeptic, or prepare to pitch a proposal internally. Produces a persuasion brief — the audience's current view and what moves them, the core argument, the proof, objection handling, the emotional and logical appeals, and the ask.
.claude/skills/mohitagw15856-persuasion-brief/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 16% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -2% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 31% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -14% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 10% | 0% |
Persuasion isn't about the strength of your logic — it's about meeting the other person where they are and giving them reasons that matter to them. This skill builds the case to win someone over: it starts from their current belief and motivations, then assembles the argument, proof, and framing most likely to move them — combining the logical case with the human one, and handling the real objection.
Ask for these only if they aren't already provided:
1. Their starting point — where they stand now and why (their incentives, constraints, prior position). You move people from where they are, not from where you wish they were.
2. The core argument — the single most compelling reason for them (not the reason that persuades you). One sentence they'd repeat to their own boss.
3. The proof — the 2–3 strongest pieces of evidence, ordered for this audience (a data person needs numbers; a relationship person needs a peer example / social proof).
4. Logic + emotion — the rational case (cost/benefit, risk reduction) and the human one (what they gain, avoid, or become). Decisions are made on both; brief both.
5. Objection handling — the real objection (name the unspoken one), and how to defuse it — ideally by addressing it before they raise it.
6. The ask & the easy yes — exactly what you're requesting, and how to lower the cost of agreeing (a pilot, a reversible step, a small first commitment).
Ethics note — persuade with true reasons that serve them too; manipulation wins once and costs the relationship.
Influence & persuasion practice — Cialdini's principles, Aristotle's ethos/pathos/logos, and audience-first framing.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 23,706 | 20,673 | -13% | 1 | 1 | 0% | 2,559 | 3,060 | +20% | 0 | 0 | — |
case-02 | fail→pass | 16,323 | 19,096 | +17% | 1 | 1 | 0% | 2,683 | 3,104 | +16% | 0 | 0 | — |
case-03 | fail→fail | 20,598 | 17,053 | -17% | 1 | 1 | 0% | 2,374 | 2,649 | +12% | 0 | 0 | — |
case-04 | pass→fail | 21,107 | 18,960 | -10% | 1 | 1 | 0% | 2,553 | 2,989 | +17% | 0 | 0 | — |
case-05 | fail→pass | 16,284 | 16,088 | -1% | 1 | 1 | 0% | 2,533 | 2,491 | -2% | 0 | 0 | — |
case-06 | fail→fail | 33,919 | 23,094 | -32% | 1 | 1 | 0% | 3,882 | 3,293 | -15% | 0 | 0 | — |
case-07 | fail→pass | 20,755 | 18,669 | -10% | 1 | 1 | 0% | 2,413 | 3,154 | +31% | 0 | 0 | — |
case-08 | fail→fail | 21,816 | 18,582 | -15% | 1 | 1 | 0% | 2,684 | 2,947 | +10% | 0 | 0 | — |
case-09 | fail→pass | 24,968 | 14,803 | -41% | 1 | 1 | 0% | 3,203 | 2,746 | -14% | 0 | 0 | — |
case-10 | pass→pass | 15,873 | 21,121 | +33% | 1 | 1 | 0% | 2,002 | 2,746 | +37% | 0 | 0 | — |
case-11 | fail→pass | 21,219 | 15,613 | -26% | 1 | 1 | 0% | 2,527 | 2,789 | +10% | 0 | 0 | — |
case-12 | fail→pass | 23,599 | 18,764 | -20% | 1 | 1 | 0% | 2,626 | 2,933 | +12% | 0 | 0 | — |
case-13 | pass→pass | 22,494 | 16,928 | -25% | 1 | 1 | 0% | 2,371 | 2,621 | +11% | 0 | 0 | — |
case-14 | fail→pass | 22,248 | 17,535 | -21% | 1 | 1 | 0% | 2,328 | 3,138 | +35% | 0 | 0 | — |
case-15 | fail→fail | 20,626 | 19,163 | -7% | 1 | 1 | 0% | 2,758 | 2,654 | -4% | 0 | 0 | — |
case-16 | fail→pass | 23,171 | 19,296 | -17% | 1 | 1 | 0% | 2,549 | 2,759 | +8% | 0 | 0 | — |
case-17 | fail→fail | 23,741 | 13,650 | -43% | 1 | 1 | 0% | 2,503 | 2,762 | +10% | 0 | 0 | — |
case-18 | fail→pass | 22,397 | 20,721 | -7% | 1 | 1 | 0% | 2,585 | 2,858 | +11% | 0 | 0 | — |
case-19 | fail→pass | 17,314 | 16,190 | -6% | 1 | 1 | 0% | 2,466 | 2,467 | +0% | 0 | 0 | — |
case-20 | pass→pass | 13,141 | 22,699 | +73% | 1 | 1 | 0% | 2,177 | 3,959 | +82% | 0 | 0 | — |
case-21 | pass→fail | 33,399 | 25,857 | -23% | 1 | 1 | 0% | 3,811 | 3,010 | -21% | 0 | 0 | — |
case-22 | pass→fail | 20,919 | 33,056 | +58% | 1 | 1 | 0% | 2,592 | 5,285 | +104% | 0 | 0 | — |
case-23 | fail→fail | 20,921 | 14,819 | -29% | 1 | 1 | 0% | 2,680 | 2,559 | -5% | 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. 23 cases were attempted. The headline lift of +30 percentage points is the difference between those two pass rates over the 23 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.