Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree. Use when user wants to stress-test a plan, get grilled on their design, or mentions "grill me".
.claude/skills/camunda-grill-me/SKILL.md| Model | Eval pass | Runs |
|---|---|---|
| gemini-3.6-flash | 100% | 9 |
| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-14 | ✓→✗ | ▼ Worse | -87% | 0% |
| case-20 | ✓→✗ | ▼ Worse | -88% | 0% |
| case-10 | ✓→✗ | ▼ Worse | -86% | 0% |
| case-16 | ✓→✗ | ▼ Worse | -90% | 0% |
| case-17 | ✓→✗ | ▼ Worse | -92% | 0% |
Interview me relentlessly about every aspect of this plan until we reach a shared understanding. Walk down each branch of the design tree, resolving dependencies between decisions one-by-one. For each question, provide your recommended answer.
If a question can be answered by exploring the codebase, explore the codebase instead.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-14 | pass→fail | 12,543 | 1,968 | -84% | 1 | 1 | 0% | 2,191 | 277 | -87% | 0 | 0 | — |
case-20 | pass→fail | 19,465 | 4,969 | -74% | 1 | 1 | 0% | 2,423 | 286 | -88% | 0 | 0 | — |
case-01 | fail→fail | 10,914 | 3,155 | -71% | 1 | 1 | 0% | 1,888 | 262 | -86% | 0 | 0 | — |
case-02 | fail→fail | 10,242 | 3,925 | -62% | 1 | 1 | 0% | 1,779 | 416 | -77% | 0 | 0 | — |
case-03 | fail→fail | 12,096 | 1,706 | -86% | 1 | 1 | 0% | 2,376 | 279 | -88% | 0 | 0 | — |
case-04 | fail→fail | 8,288 | 3,395 | -59% | 1 | 1 | 0% | 1,509 | 276 | -82% | 0 | 0 | — |
case-05 | pass→pass | 15,814 | 8,491 | -46% | 1 | 1 | 0% | 2,734 | 1,425 | -48% | 0 | 0 | — |
case-06 | fail→fail | 7,797 | 2,941 | -62% | 1 | 1 | 0% | 1,568 | 279 | -82% | 0 | 0 | — |
case-07 | fail→fail | 10,545 | 3,795 | -64% | 1 | 1 | 0% | 1,819 | 243 | -87% | 0 | 0 | — |
case-08 | fail→fail | 14,119 | 2,529 | -82% | 1 | 1 | 0% | 2,636 | 233 | -91% | 0 | 0 | — |
case-09 | fail→fail | 11,831 | 4,009 | -66% | 1 | 1 | 0% | 2,329 | 297 | -87% | 0 | 0 | — |
case-10 | pass→fail | 10,040 | 2,701 | -73% | 1 | 1 | 0% | 1,752 | 252 | -86% | 0 | 0 | — |
case-11 | fail→fail | 13,660 | 2,674 | -80% | 1 | 1 | 0% | 2,558 | 261 | -90% | 0 | 0 | — |
case-12 | fail→fail | 13,019 | 2,731 | -79% | 1 | 1 | 0% | 2,258 | 166 | -93% | 0 | 0 | — |
case-13 | fail→fail | 14,185 | 2,426 | -83% | 1 | 1 | 0% | 2,549 | 226 | -91% | 0 | 0 | — |
case-15 | fail→fail | 11,750 | 2,717 | -77% | 1 | 1 | 0% | 2,130 | 229 | -89% | 0 | 0 | — |
case-16 | pass→fail | 13,947 | 1,682 | -88% | 1 | 1 | 0% | 2,451 | 255 | -90% | 0 | 0 | — |
case-17 | pass→fail | 12,240 | 2,582 | -79% | 1 | 1 | 0% | 2,484 | 205 | -92% | 0 | 0 | — |
case-18 | pass→fail | 12,520 | 5,146 | -59% | 1 | 1 | 0% | 2,466 | 245 | -90% | 0 | 0 | — |
case-19 | fail→fail | 8,834 | 3,012 | -66% | 1 | 1 | 0% | 1,503 | 245 | -84% | 0 | 0 | — |
case-21 | fail→fail | 7,575 | 3,869 | -49% | 1 | 1 | 0% | 1,439 | 251 | -83% | 0 | 0 | — |
case-22 | pass→fail | 10,465 | 3,066 | -71% | 1 | 1 | 0% | 2,642 | 315 | -88% | 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, and 9 counted toward the lift figure. The other 13 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of -50 percentage points is the difference between those two pass rates over the 9 comparable cases. 11 cases got worse with the skill loaded, and they are included in that figure.
Other measured skills in the registry, with their headline benchmark lift.