Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Decline a request, push back on scope, or protect priorities without burning the relationship. Use when asked how to say no, turn down a request, push back on your boss/stakeholder, decline extra work, or protect the roadmap from a pet feature. Produces a graceful, firm response — the no, the honest why, an alternative or trade-off, and the exact wording, tuned to who's asking.
.claude/skills/mohitagw15856-saying-no/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-20 | ✓→✗ | ▼ Worse | 143% | 0% |
| case-01 | ✓→✓ | = Same ✓ | 38% | 0% |
| case-02 | ✓→✓ | = Same ✓ | 25% | 0% |
| case-03 | ✓→✓ | = Same ✓ | 64% | 0% |
| case-04 | ✓→✓ | = Same ✓ | 39% | 0% |
Most people say yes to things they should decline because they don't know how to say no without seeming difficult — and then over-commit, miss what matters, or resent it. A good no is clear, respectful, and offers a path: it declines the request while honouring the relationship and, often, reframes it as a trade-off rather than a flat refusal. This skill writes that no.
Ask for these only if they aren't already provided:
1. The frame — is this a flat no, a "not now," a "yes if trade-off]," or a "no, but here's another way"? Pick the honest one. Most good nos are trade-offs, not refusals.
2. The response — the actual wording, structured:
3. For "no" to a boss / stakeholder — frame it as protecting their goal: surface the trade-off and let them choose ("I can take this on, but the launch slips a week — your call"). This makes the cost visible without insubordination.
4. Hold the line — a prepared response if they push back, so you don't cave into a reluctant yes.
Tone note — warm and firm; brief beats over-justified (a pile of reasons invites negotiation of each).
Boundary-setting and negotiation practice — the "positive no" (William Ury), trade-off framing, and protecting priorities.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 17,935 | 16,198 | -10% | 1 | 1 | 0% | 1,866 | 2,574 | +38% | 0 | 0 | — |
case-02 | pass→pass | 22,188 | 19,047 | -14% | 1 | 1 | 0% | 2,377 | 2,978 | +25% | 0 | 0 | — |
case-03 | pass→pass | 15,284 | 11,868 | -22% | 1 | 1 | 0% | 1,603 | 2,624 | +64% | 0 | 0 | — |
case-04 | pass→pass | 17,718 | 16,737 | -6% | 1 | 1 | 0% | 1,711 | 2,373 | +39% | 0 | 0 | — |
case-05 | pass→pass | 13,367 | 16,213 | +21% | 1 | 1 | 0% | 1,659 | 2,488 | +50% | 0 | 0 | — |
case-06 | pass→pass | 15,961 | 17,013 | +7% | 1 | 1 | 0% | 1,701 | 2,451 | +44% | 0 | 0 | — |
case-07 | pass→pass | 12,117 | 14,058 | +16% | 1 | 1 | 0% | 1,121 | 1,945 | +74% | 0 | 0 | — |
case-08 | pass→pass | 18,514 | 8,656 | -53% | 1 | 1 | 0% | 1,860 | 2,087 | +12% | 0 | 0 | — |
case-09 | pass→pass | 12,264 | 16,932 | +38% | 1 | 1 | 0% | 1,669 | 2,592 | +55% | 0 | 0 | — |
case-10 | pass→pass | 13,599 | 8,756 | -36% | 1 | 1 | 0% | 1,182 | 2,114 | +79% | 0 | 0 | — |
case-11 | pass→pass | 16,223 | 13,073 | -19% | 1 | 1 | 0% | 1,529 | 1,999 | +31% | 0 | 0 | — |
case-12 | fail→fail | 18,327 | 10,995 | -40% | 1 | 1 | 0% | 1,899 | 2,553 | +34% | 0 | 0 | — |
case-13 | pass→pass | 12,080 | 12,835 | +6% | 1 | 1 | 0% | 1,616 | 1,812 | +12% | 0 | 0 | — |
case-14 | pass→pass | 14,253 | 17,723 | +24% | 1 | 1 | 0% | 1,294 | 2,512 | +94% | 0 | 0 | — |
case-15 | pass→pass | 12,851 | 14,947 | +16% | 1 | 1 | 0% | 1,241 | 2,031 | +64% | 0 | 0 | — |
case-16 | pass→pass | 16,787 | 17,847 | +6% | 1 | 1 | 0% | 1,676 | 2,566 | +53% | 0 | 0 | — |
case-17 | pass→pass | 13,163 | 8,816 | -33% | 1 | 1 | 0% | 1,346 | 2,121 | +58% | 0 | 0 | — |
case-18 | pass→pass | 9,623 | 18,719 | +95% | 1 | 1 | 0% | 1,473 | 2,545 | +73% | 0 | 0 | — |
case-19 | pass→pass | 17,014 | 17,072 | +0% | 1 | 1 | 0% | 1,675 | 2,537 | +51% | 0 | 0 | — |
case-20 | pass→fail | 21,092 | 32,808 | +56% | 1 | 1 | 0% | 2,160 | 5,241 | +143% | 0 | 0 | — |
case-21 | pass→pass | 17,096 | 27,468 | +61% | 1 | 1 | 0% | 1,717 | 4,064 | +137% | 0 | 0 | — |
case-22 | pass→pass | 21,566 | 22,891 | +6% | 1 | 1 | 0% | 2,477 | 3,589 | +45% | 0 | 0 | — |
case-23 | pass→pass | 14,392 | 18,589 | +29% | 1 | 1 | 0% | 1,281 | 2,514 | +96% | 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 -100 percentage points is the difference between those two pass rates over the 23 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Other measured skills in the registry, with their headline benchmark lift.