Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Expert guide for Backend.AI distributed computing platform. Automatically activates when users ask about: - Backend.AI architecture, components (Manager, Agent, Storage Proxy, Webserver, App Proxy) - Features (session scheduling, Sokovan orchestrator, multi-tenancy, resource allocation) - APIs (REST, GraphQL), authentication, RBAC authorization - Container runtime (kernels, jail sandbox, hook library, virtual folders) - Accelerator support (CUDA, ROCm, TPU, NPU, Graphcore IPU) - Client SD
.claude/skills/microck-backend-ai-guide/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 27% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 37% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 23% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 90% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 78% | 0% |
This skill provides expert-level information about the Backend.AI platform by:
Activate this skill when the user asks about:
src/ai/backend/manager/README.mdsrc/ai/backend/agent/README.mdsrc/ai/backend/storage/README.mdsrc/ai/backend/web/README.mdsrc/ai/backend/appproxy/README.mdArchitecture Questions
Component Questions
Integration Questions
Setup Questions
Structure answers as follows:
markdown## [Direct Answer to Question] [Concise, direct answer based on official documentation] ## Details [Supporting information from fetched documentation] ### Component Interactions (if applicable) [How different components work together] ## Technical Specifications (if applicable) - Requirements: [versions, dependencies] - Configuration: [relevant settings] - APIs: [REST/GraphQL endpoints] ## Source Documentation - Main: [URL to main README] - Component: [URLs to component READMEs] - Additional: [URLs to other relevant docs]
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-11 | pass→pass | 10,005 | 4,837 | -52% | 1 | 1 | 0% | 1,409 | 2,016 | +43% | 0 | 0 | — |
case-01 | fail→fail | 23,416 | 18,794 | -20% | 1 | 1 | 0% | 3,856 | 3,756 | -3% | 0 | 0 | — |
case-02 | fail→fail | 20,582 | 12,868 | -37% | 1 | 1 | 0% | 3,583 | 3,372 | -6% | 0 | 0 | — |
case-03 | pass→fail | 15,459 | 6,947 | -55% | 1 | 1 | 0% | 2,604 | 1,649 | -37% | 0 | 0 | — |
case-04 | fail→pass | 9,949 | 7,770 | -22% | 1 | 1 | 0% | 1,656 | 2,101 | +27% | 0 | 0 | — |
case-10 | fail→pass | 18,237 | 6,528 | -64% | 1 | 1 | 0% | 1,673 | 2,287 | +37% | 0 | 0 | — |
case-05 | fail→pass | 10,219 | 9,194 | -10% | 1 | 1 | 0% | 1,793 | 2,201 | +23% | 0 | 0 | — |
case-06 | pass→fail | 4,378 | 5,768 | +32% | 1 | 1 | 0% | 722 | 1,749 | +142% | 0 | 0 | — |
case-07 | fail→pass | 6,728 | 3,712 | -45% | 1 | 1 | 0% | 1,086 | 2,058 | +90% | 0 | 0 | — |
case-08 | fail→pass | 7,956 | 5,867 | -26% | 1 | 1 | 0% | 1,290 | 2,294 | +78% | 0 | 0 | — |
case-09 | pass→pass | 7,946 | 7,440 | -6% | 1 | 1 | 0% | 1,218 | 2,650 | +118% | 0 | 0 | — |
case-12 | fail→pass | 3,630 | 2,362 | -35% | 1 | 1 | 0% | 611 | 1,679 | +175% | 0 | 0 | — |
case-13 | pass→pass | 9,699 | 10,369 | +7% | 1 | 1 | 0% | 1,601 | 2,444 | +53% | 0 | 0 | — |
case-14 | fail→fail | 15,935 | 7,831 | -51% | 1 | 1 | 0% | 2,527 | 2,664 | +5% | 0 | 0 | — |
case-15 | pass→pass | 4,082 | 3,497 | -14% | 1 | 1 | 0% | 530 | 1,880 | +255% | 0 | 0 | — |
case-16 | fail→pass | 14,327 | 27,603 | +93% | 1 | 1 | 0% | 2,289 | 5,675 | +148% | 0 | 0 | — |
case-17 | pass→pass | 17,409 | 17,399 | -0% | 1 | 1 | 0% | 2,792 | 4,328 | +55% | 0 | 0 | — |
case-18 | pass→pass | 15,283 | 14,465 | -5% | 1 | 1 | 0% | 2,363 | 3,521 | +49% | 0 | 0 | — |
case-19 | pass→pass | 12,063 | 4,918 | -59% | 1 | 1 | 0% | 1,696 | 2,021 | +19% | 0 | 0 | — |
case-20 | pass→pass | 13,964 | 14,513 | +4% | 1 | 1 | 0% | 2,758 | 4,119 | +49% | 0 | 0 | — |
case-21 | pass→pass | 8,091 | 6,637 | -18% | 1 | 1 | 0% | 1,327 | 2,527 | +90% | 0 | 0 | — |
case-22 | pass→pass | 9,443 | 9,598 | +2% | 1 | 1 | 0% | 1,718 | 3,101 | +81% | 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 20 counted toward the lift figure. The other 2 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 +23 percentage points is the difference between those two pass rates over the 20 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.