Install any skill in seconds. Free to start, no credit card required.
Get Started Free →When the user needs to evaluate competitors, understand the competitive landscape, or position their product against alternatives.
.claude/skills/mkurman-competitive-analysis/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 35% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -28% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 152% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 45% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 167% | 0% |
| | Company Profile | Background, founding, funding, team size, headquarters | | Core Product Strengths | Key capabilities, UX quality, integrations, technical advantages | | Product Gaps | Missing features, weak areas, customer complaints | | Business Model & Pricing | Model type, price points, free tier, enterprise options | | Competitive Threats | What makes them dangerous -- momentum, funding, distribution, brand |
market-research -- Pair competitive analysis with market sizing to understand both the landscape and the size of the prize.prd-writing -- Use competitive gaps to inform the solution section of a PRD. Build what competitors miss.user-research-synthesis -- Ground feature comparisons in what customers actually value, not what competitors assume they value.User: "I need a competitive landscape slide for our Series A deck."
Good output: A market overview paragraph, profiles of 4-5 key competitors highlighting their weaknesses relative to the startup's strengths, a differentiation map showing clear whitespace where the startup sits, and a positioning recommendation that tells a narrative about why competitive dynamics favor the startup's approach.
User: "We keep losing deals to Competitor X. Help us understand why and how to win."
Good output: Deep analysis of Competitor X's strengths in the dimensions that matter to buyers (e.g., enterprise readiness, integrations, brand trust), identification of the specific evaluation criteria where they win, and actionable recommendations to either match their advantages on must-have dimensions or shift the evaluation criteria to dimensions where you are stronger.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 30,232 | 30,740 | +2% | 1 | 1 | 0% | 4,570 | 5,241 | +15% | 0 | 0 | — |
case-02 | fail→pass | 26,846 | 32,539 | +21% | 1 | 1 | 0% | 4,114 | 5,537 | +35% | 0 | 0 | — |
case-03 | pass→pass | 20,860 | 22,111 | +6% | 1 | 1 | 0% | 3,549 | 4,532 | +28% | 0 | 0 | — |
case-04 | pass→pass | 25,819 | 33,172 | +28% | 1 | 1 | 0% | 4,235 | 6,152 | +45% | 0 | 0 | — |
case-05 | pass→pass | 15,122 | 21,878 | +45% | 1 | 1 | 0% | 2,497 | 4,192 | +68% | 0 | 0 | — |
case-06 | fail→fail | 18,066 | 22,260 | +23% | 1 | 1 | 0% | 2,623 | 4,097 | +56% | 0 | 0 | — |
case-07 | fail→fail | 4,093 | 18,894 | +362% | 1 | 1 | 0% | 605 | 3,504 | +479% | 0 | 0 | — |
case-08 | fail→pass | 25,220 | 13,887 | -45% | 1 | 1 | 0% | 4,043 | 2,916 | -28% | 0 | 0 | — |
case-09 | fail→pass | 5,561 | 9,050 | +63% | 1 | 1 | 0% | 920 | 2,318 | +152% | 0 | 0 | — |
case-10 | fail→pass | 18,904 | 19,719 | +4% | 1 | 1 | 0% | 2,503 | 3,637 | +45% | 0 | 0 | — |
case-16 | pass→pass | 16,830 | 22,532 | +34% | 1 | 1 | 0% | 2,430 | 3,859 | +59% | 0 | 0 | — |
case-11 | fail→pass | 8,601 | 17,188 | +100% | 1 | 1 | 0% | 1,205 | 3,219 | +167% | 0 | 0 | — |
case-12 | fail→pass | 15,299 | 13,778 | -10% | 1 | 1 | 0% | 2,175 | 2,905 | +34% | 0 | 0 | — |
case-13 | fail→fail | 15,613 | 17,529 | +12% | 1 | 1 | 0% | 2,130 | 3,185 | +50% | 0 | 0 | — |
case-14 | pass→pass | 18,093 | 19,330 | +7% | 1 | 1 | 0% | 2,561 | 3,583 | +40% | 0 | 0 | — |
case-15 | pass→fail | 18,901 | 20,381 | +8% | 1 | 1 | 0% | 2,676 | 3,766 | +41% | 0 | 0 | — |
case-17 | pass→pass | 19,673 | 17,329 | -12% | 1 | 1 | 0% | 2,942 | 3,342 | +14% | 0 | 0 | — |
case-18 | fail→pass | 17,694 | 19,375 | +10% | 1 | 1 | 0% | 2,806 | 3,664 | +31% | 0 | 0 | — |
case-19 | fail→fail | 18,500 | 20,634 | +12% | 1 | 1 | 0% | 2,739 | 3,956 | +44% | 0 | 0 | — |
case-20 | fail→fail | 16,250 | 19,023 | +17% | 1 | 1 | 0% | 2,413 | 3,508 | +45% | 0 | 0 | — |
case-21 | pass→pass | 16,736 | 16,467 | -2% | 1 | 1 | 0% | 2,409 | 3,157 | +31% | 0 | 0 | — |
case-22 | pass→pass | 15,853 | 15,990 | +1% | 1 | 1 | 0% | 2,287 | 3,114 | +36% | 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. 1 case got worse with the skill loaded, and it is 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.