Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →/close-plan <deal_name> [--target-date <date>]
.claude/skills/miosa-osa-close-plan/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -31% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -29% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -70% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -36% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -68% | 0% |
/close-plan <deal_name> --target-date <date>]
Generate a mutual action plan (MAP) with customer-side milestones, internal milestones, risk assessment, and timeline to close.
| Arg | Type | Required | Description | |-----|------|----------|-------------| | deal_name | string | Yes | Name of the deal | | --target-date | date | No | Target close date (ISO 8601). If omitted, estimates from deal context. |
Genre: plan Format: Markdown (Mutual Action Plan template)
Produces:
1. Pull MEDDPICC scorecard for the deal
2. Verify all 8 letters score >= 2 (prerequisite for close plan)
3. If any letter < 2: return gap warning, recommend /qualify first
4. Map decision process milestones backward from target close date
5. Add internal milestones (proposal, legal review, contract generation)
6. Identify risks from MEDDPICC gaps and deal history
7. Generate negotiation guardrails per discount authority matrix
8. Output complete MAP per closer's template/close-plan "Acme Corp Enterprise Deal"
/close-plan "Acme Corp" --target-date 2026-06-30| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-08 | fail→fail | 14,081 | 3,216 | -77% | 1 | 1 | 0% | 1,772 | 775 | -56% | 0 | 0 | — |
case-01 | fail→pass | 28,819 | 20,303 | -30% | 1 | 1 | 0% | 5,345 | 3,675 | -31% | 0 | 0 | — |
case-02 | fail→fail | 23,712 | 17,986 | -24% | 1 | 1 | 0% | 3,445 | 3,351 | -3% | 0 | 0 | — |
case-03 | fail→pass | 22,912 | 15,863 | -31% | 1 | 1 | 0% | 3,735 | 2,639 | -29% | 0 | 0 | — |
case-04 | fail→fail | 16,057 | 17,844 | +11% | 1 | 1 | 0% | 2,207 | 3,435 | +56% | 0 | 0 | — |
case-05 | pass→pass | 13,677 | 5,724 | -58% | 1 | 1 | 0% | 1,973 | 1,406 | -29% | 0 | 0 | — |
case-06 | pass→pass | 13,228 | 12,096 | -9% | 1 | 1 | 0% | 2,082 | 1,949 | -6% | 0 | 0 | — |
case-07 | pass→pass | 11,258 | 11,149 | -1% | 1 | 1 | 0% | 1,668 | 1,769 | +6% | 0 | 0 | — |
case-09 | pass→pass | 9,940 | 3,425 | -66% | 1 | 1 | 0% | 1,414 | 934 | -34% | 0 | 0 | — |
case-10 | pass→pass | 14,230 | 10,691 | -25% | 1 | 1 | 0% | 2,203 | 1,990 | -10% | 0 | 0 | — |
case-11 | fail→pass | 19,002 | 2,167 | -89% | 1 | 1 | 0% | 2,552 | 767 | -70% | 0 | 0 | — |
case-12 | fail→pass | 13,312 | 4,304 | -68% | 1 | 1 | 0% | 1,635 | 1,050 | -36% | 0 | 0 | — |
case-13 | pass→pass | 8,904 | 2,516 | -72% | 1 | 1 | 0% | 1,539 | 875 | -43% | 0 | 0 | — |
case-14 | pass→pass | 14,271 | 8,726 | -39% | 1 | 1 | 0% | 2,181 | 1,658 | -24% | 0 | 0 | — |
case-15 | fail→pass | 15,081 | 2,207 | -85% | 1 | 1 | 0% | 2,424 | 778 | -68% | 0 | 0 | — |
case-16 | pass→pass | 9,504 | 2,705 | -72% | 1 | 1 | 0% | 1,174 | 751 | -36% | 0 | 0 | — |
case-17 | fail→pass | 6,406 | 1,878 | -71% | 1 | 1 | 0% | 885 | 673 | -24% | 0 | 0 | — |
case-18 | fail→pass | 15,101 | 7,848 | -48% | 1 | 1 | 0% | 1,702 | 1,477 | -13% | 0 | 0 | — |
case-19 | fail→pass | 8,943 | 1,996 | -78% | 1 | 1 | 0% | 1,353 | 665 | -51% | 0 | 0 | — |
case-20 | fail→fail | 5,198 | 6,369 | +23% | 1 | 1 | 0% | 659 | 1,185 | +80% | 0 | 0 | — |
case-21 | fail→fail | 46,269 | 19,060 | -59% | 1 | 1 | 0% | 6,651 | 3,894 | -41% | 0 | 0 | — |
case-22 | fail→fail | 10,377 | 12,844 | +24% | 1 | 1 | 0% | 1,682 | 2,472 | +47% | 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 +36 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.