Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Cross-role engineering grill — Matt Pocock 7 questions per role × 3 roles (fullstack / frontend / backend) = up to 21 forcing questions, one per turn, with canon citations and kill criteria. Default: ask which lane first; `--all` runs all 21.
.claude/skills/alirezarezvani-csengineer-grill-cross-role-engineering-forcing-question-grill/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 18% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 208% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 115% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 12% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -7% | 0% |
Walk the user through the Matt Pocock forcing-question discipline before they lock any engineering decision. This is the grill-with-docs pattern (canon-anchored, recommended answers, kill criteria) applied across the three engineering role lanes.
$ARGUMENTS
--lane <name> is supplied: walk only that lane's 7 questions.--lane all: ask the user: "Fullstack (7 Qs about team / stack / scale), Frontend (7 Qs about device / rendering / bundle / a11y), or Backend (7 Qs about QPS / tenancy / pattern / SLO)? Or all for all 21."Questions live in engineering-team/skills/senior-fullstack/references/forcing_questions.md. Summary:
Questions live in engineering-team/skills/senior-frontend/references/forcing_questions.md. Summary:
Questions live in engineering-team/skills/senior-backend/references/forcing_questions.md. Summary:
engineering/grill-me)/tmp/engineer-grill-<lane>-<date>.md so the conversation survives compaction.python engineering-team/skills/senior-fullstack/scripts/fullstack_decision_engine.py ...python engineering-team/skills/senior-frontend/scripts/frontend_decision_engine.py ...python engineering-team/skills/senior-backend/scripts/backend_decision_engine.py .../tmp/engineer-grill-<lane>-<date>.md./cs:fullstack-review, /cs:frontend-review, /cs:backend-review — single-lane deep dives/karpathy-check — Karpathy review before commit/cs:grill-bizops, /cs:grill-commercial — sibling cross-domain grills (BizOps + Commercial v2.8.0)| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 11,972 | 3,724 | -69% | 1 | 1 | 0% | 1,890 | 1,908 | +1% | 0 | 0 | — |
case-02 | fail→fail | 22,056 | 3,941 | -82% | 1 | 1 | 0% | 3,779 | 1,939 | -49% | 0 | 0 | — |
case-03 | fail→fail | 28,735 | 4,515 | -84% | 1 | 1 | 0% | 3,862 | 1,983 | -49% | 0 | 0 | — |
case-04 | fail→pass | 8,565 | 3,225 | -62% | 1 | 1 | 0% | 1,449 | 1,705 | +18% | 0 | 0 | — |
case-05 | fail→pass | 16,394 | 3,393 | -79% | 1 | 1 | 0% | 556 | 1,710 | +208% | 0 | 0 | — |
case-06 | fail→pass | 5,405 | 3,308 | -39% | 1 | 1 | 0% | 818 | 1,757 | +115% | 0 | 0 | — |
case-07 | fail→pass | 9,006 | 3,044 | -66% | 1 | 1 | 0% | 1,525 | 1,712 | +12% | 0 | 0 | — |
case-08 | fail→fail | 15,631 | 3,324 | -79% | 1 | 1 | 0% | 3,201 | 1,839 | -43% | 0 | 0 | — |
case-09 | fail→fail | 19,409 | 3,343 | -83% | 1 | 1 | 0% | 3,393 | 1,835 | -46% | 0 | 0 | — |
case-10 | fail→fail | 20,803 | 3,565 | -83% | 1 | 1 | 0% | 3,699 | 1,782 | -52% | 0 | 0 | — |
case-11 | fail→pass | 11,390 | 3,879 | -66% | 1 | 1 | 0% | 1,980 | 1,843 | -7% | 0 | 0 | — |
case-12 | pass→pass | 12,013 | 7,605 | -37% | 1 | 1 | 0% | 2,008 | 2,481 | +24% | 0 | 0 | — |
case-13 | fail→fail | 13,919 | 2,198 | -84% | 1 | 1 | 0% | 2,366 | 1,552 | -34% | 0 | 0 | — |
case-14 | fail→pass | 13,212 | 8,470 | -36% | 1 | 1 | 0% | 2,308 | 2,648 | +15% | 0 | 0 | — |
case-15 | pass→pass | 4,664 | 3,370 | -28% | 1 | 1 | 0% | 828 | 1,755 | +112% | 0 | 0 | — |
case-16 | fail→fail | 6,166 | 3,968 | -36% | 1 | 1 | 0% | 965 | 1,872 | +94% | 0 | 0 | — |
case-17 | fail→pass | 5,990 | 3,468 | -42% | 1 | 1 | 0% | 889 | 1,751 | +97% | 0 | 0 | — |
case-18 | fail→pass | 6,038 | 3,561 | -41% | 1 | 1 | 0% | 1,074 | 1,778 | +66% | 0 | 0 | — |
case-19 | fail→pass | 5,670 | 3,372 | -41% | 1 | 1 | 0% | 928 | 1,782 | +92% | 0 | 0 | — |
case-20 | pass→pass | 3,194 | 4,739 | +48% | 1 | 1 | 0% | 592 | 2,034 | +244% | 0 | 0 | — |
case-21 | pass→fail | 2,735 | 7,912 | +189% | 1 | 1 | 0% | 514 | 2,722 | +430% | 0 | 0 | — |
case-22 | pass→fail | 6,560 | 7,508 | +14% | 1 | 1 | 0% | 1,177 | 2,428 | +106% | 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 21 counted toward the lift figure. The other 1 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 +32 percentage points is the difference between those two pass rates over the 21 comparable cases. 2 cases got worse with the skill loaded, and they are 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.