Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →/battlecard <competitor_name> [--refresh]
.claude/skills/miosa-osa-battlecard/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 39% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 50% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -37% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -37% | 0% |
/battlecard <competitor_name> --refresh]
Generate or update a competitive battlecard with positioning, strengths/weaknesses, pricing intelligence, landmine questions, and objection responses.
| Arg | Type | Required | Description | |-----|------|----------|-------------| | competitor_name | string | Yes | Competitor to analyze | | --refresh | flag | No | Force refresh even if battlecard exists and is recent |
Genre: battlecard Format: Markdown (Competitive Battlecard template)
Produces:
1. Check if existing battlecard exists and is < 30 days old (skip research if current)
2. If --refresh or stale: researcher gathers fresh competitive intel
3. Verify all claims against independent sources
4. Rate every data point with source reliability (A-E)
5. Copywriter crafts objection responses and landmine question phrasing
6. Output complete battlecard per researcher's template/battlecard "Competitor X"
/battlecard "Competitor X" --refresh| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 19,644 | 19,427 | -1% | 1 | 1 | 0% | 2,955 | 3,283 | +11% | 0 | 0 | — |
case-02 | fail→pass | 17,293 | 30,172 | +74% | 1 | 1 | 0% | 2,419 | 3,357 | +39% | 0 | 0 | — |
case-03 | fail→pass | 24,725 | 33,573 | +36% | 1 | 1 | 0% | 3,109 | 4,671 | +50% | 0 | 0 | — |
case-04 | fail→pass | 11,640 | 6,209 | -47% | 1 | 1 | 0% | 1,940 | 1,221 | -37% | 0 | 0 | — |
case-05 | pass→pass | 6,530 | 3,650 | -44% | 1 | 1 | 0% | 988 | 1,003 | +2% | 0 | 0 | — |
case-06 | pass→pass | 7,620 | 4,756 | -38% | 1 | 1 | 0% | 1,180 | 1,091 | -8% | 0 | 0 | — |
case-07 | pass→pass | 13,347 | 5,263 | -61% | 1 | 1 | 0% | 2,005 | 1,172 | -42% | 0 | 0 | — |
case-08 | fail→pass | 16,094 | 8,820 | -45% | 1 | 1 | 0% | 2,746 | 1,724 | -37% | 0 | 0 | — |
case-09 | fail→pass | 24,240 | 32,103 | +32% | 1 | 1 | 0% | 3,408 | 3,649 | +7% | 0 | 0 | — |
case-10 | pass→pass | 17,174 | 17,117 | -0% | 1 | 1 | 0% | 2,437 | 3,078 | +26% | 0 | 0 | — |
case-11 | fail→pass | 10,623 | 2,210 | -79% | 1 | 1 | 0% | 1,295 | 714 | -45% | 0 | 0 | — |
case-16 | fail→pass | 8,422 | 1,813 | -78% | 1 | 1 | 0% | 1,178 | 656 | -44% | 0 | 0 | — |
case-12 | pass→fail | 12,843 | 3,027 | -76% | 1 | 1 | 0% | 1,927 | 928 | -52% | 0 | 0 | — |
case-13 | fail→pass | 12,446 | 9,443 | -24% | 1 | 1 | 0% | 1,863 | 1,590 | -15% | 0 | 0 | — |
case-14 | fail→pass | 9,865 | 3,175 | -68% | 1 | 1 | 0% | 1,355 | 859 | -37% | 0 | 0 | — |
case-15 | pass→pass | 14,701 | 16,089 | +9% | 1 | 1 | 0% | 2,109 | 2,759 | +31% | 0 | 0 | — |
case-17 | pass→pass | 28,794 | 23,031 | -20% | 1 | 1 | 0% | 1,985 | 1,616 | -19% | 0 | 0 | — |
case-18 | pass→pass | 11,452 | 4,560 | -60% | 1 | 1 | 0% | 1,828 | 1,179 | -36% | 0 | 0 | — |
case-19 | pass→pass | 8,844 | 2,468 | -72% | 1 | 1 | 0% | 1,275 | 656 | -49% | 0 | 0 | — |
case-20 | pass→pass | 17,840 | 13,221 | -26% | 1 | 1 | 0% | 2,222 | 2,366 | +6% | 0 | 0 | — |
case-21 | fail→pass | 13,616 | 12,139 | -11% | 1 | 1 | 0% | 2,072 | 2,203 | +6% | 0 | 0 | — |
case-22 | fail→fail | 8,608 | 2,858 | -67% | 1 | 1 | 0% | 819 | 749 | -9% | 0 | 0 | — |
case-23 | pass→pass | 17,179 | 15,988 | -7% | 1 | 1 | 0% | 3,063 | 2,872 | -6% | 0 | 0 | — |
case-24 | pass→pass | 19,783 | 18,316 | -7% | 1 | 1 | 0% | 2,565 | 2,729 | +6% | 0 | 0 | — |
case-25 | pass→pass | 17,203 | 19,848 | +15% | 1 | 1 | 0% | 3,513 | 3,588 | +2% | 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. 25 cases were attempted. The headline lift of +40 percentage points is the difference between those two pass rates over the 25 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.