Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Architecture knowledge reference covering API design, security architecture, cloud-native patterns, caching strategies, message queues, and data security. Use when designing system architecture, APIs, or cloud-native infrastructure.
.claude/skills/telagod-designing-architectures/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-09 | ✗→✓ | ▲ Improved | 31% | 0% |
| case-21 | ✓→✗ | ▼ Worse | 45% | 0% |
| case-14 | ✓→✓ | ＝ Same ✓ | 11% | 0% |
| case-12 | ✓→✓ | ＝ Same ✓ | 25% | 0% |
| case-13 | ✓→✓ | ＝ Same ✓ | 14% | 0% |
> 判断先于执行：决定「是否做 / 选什么 / 如何取舍」（栈、方案、架构、权衡）前，先读领域判断内核 skills/_kernel/backend/SKILL.md——它管 judgment，本秘典管 execution；冲突时以内核判断为准。
| 意图 | 秘典 | 核心 | |------|------|------| | API 设计 | api-design | RESTful、GraphQL、gRPC、OpenAPI | | 安全架构 | security-arch | 零信任、IAM、威胁建模、合规 | | 云原生 | cloud-native | 容器、K8s、Serverless、Service Mesh | | 消息队列 | message-queue | Kafka、RabbitMQ、事件驱动、CQRS | | 缓存 | caching | Redis、CDN、一致性、穿透/雪崩 |
| 决策点 | 选项 A | 选项 B | 判据 | |--------|--------|--------|------| | 同步 vs 异步 | REST/gRPC 同步调用 | 消息队列异步 | 延迟敏感→同步；解耦/削峰→异步 | | 单体 vs 微服务 | 单体(模块化) | 微服务 | 团队<5→单体；独立部署需求→微服务 | | SQL vs NoSQL | RDBMS | MongoDB/DynamoDB | 强一致/关联→SQL；灵活 schema/高吞吐→NoSQL | | 缓存策略 | Cache-Aside | Write-Through | 读多写少→Aside；写后即读→Through | | API 风格 | REST | GraphQL | 资源型 CRUD→REST；复杂聚合/前端驱动→GraphQL | | 事件架构 | 事件通知 | 事件溯源(ES) | 简单解耦→通知；审计/回溯→ES+CQRS |
SOLID: S单一职责 O开闭 L里氏替换 I接口隔离 D依赖倒置
分布式: CAP定理 | BASE最终一致 | 幂等设计
安全: 纵深防御 | 最小权限 | 零信任
扩展: 水平优先 | 无状态服务 | 数据分片| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-14 | pass→pass | 18,464 | 16,769 | -9% | 1 | 1 | 0% | 3,117 | 3,456 | +11% | 0 | 0 | — |
case-12 | pass→pass | 27,442 | 31,526 | +15% | 1 | 1 | 0% | 4,300 | 5,394 | +25% | 0 | 0 | — |
case-13 | pass→pass | 20,050 | 19,286 | -4% | 1 | 1 | 0% | 3,443 | 3,923 | +14% | 0 | 0 | — |
case-01 | pass→pass | 26,525 | 30,085 | +13% | 1 | 1 | 0% | 4,270 | 5,333 | +25% | 0 | 0 | — |
case-02 | pass→pass | 24,696 | 24,147 | -2% | 1 | 1 | 0% | 3,806 | 4,416 | +16% | 0 | 0 | — |
case-03 | pass→pass | 23,510 | 22,020 | -6% | 1 | 1 | 0% | 3,468 | 3,904 | +13% | 0 | 0 | — |
case-04 | pass→pass | 23,037 | 22,163 | -4% | 1 | 1 | 0% | 3,534 | 3,840 | +9% | 0 | 0 | — |
case-05 | pass→pass | 24,547 | 20,704 | -16% | 1 | 1 | 0% | 3,540 | 3,627 | +2% | 0 | 0 | — |
case-06 | pass→pass | 26,811 | 29,976 | +12% | 1 | 1 | 0% | 3,995 | 4,873 | +22% | 0 | 0 | — |
case-07 | pass→pass | 23,375 | 23,159 | -1% | 1 | 1 | 0% | 3,401 | 3,788 | +11% | 0 | 0 | — |
case-08 | pass→pass | 22,926 | 25,479 | +11% | 1 | 1 | 0% | 3,460 | 4,561 | +32% | 0 | 0 | — |
case-09 | fail→pass | 23,805 | 27,433 | +15% | 1 | 1 | 0% | 3,667 | 4,794 | +31% | 0 | 0 | — |
case-10 | pass→pass | 21,379 | 18,007 | -16% | 1 | 1 | 0% | 3,572 | 3,391 | -5% | 0 | 0 | — |
case-11 | pass→pass | 19,455 | 17,665 | -9% | 1 | 1 | 0% | 3,031 | 3,578 | +18% | 0 | 0 | — |
case-15 | pass→pass | 18,441 | 21,112 | +14% | 1 | 1 | 0% | 2,642 | 3,687 | +40% | 0 | 0 | — |
case-16 | pass→pass | 21,030 | 21,994 | +5% | 1 | 1 | 0% | 3,073 | 3,931 | +28% | 0 | 0 | — |
case-17 | pass→pass | 18,496 | 20,269 | +10% | 1 | 1 | 0% | 2,755 | 3,647 | +32% | 0 | 0 | — |
case-18 | pass→pass | 25,393 | 25,840 | +2% | 1 | 1 | 0% | 3,878 | 4,603 | +19% | 0 | 0 | — |
case-19 | pass→pass | 25,576 | 29,288 | +15% | 1 | 1 | 0% | 3,788 | 5,401 | +43% | 0 | 0 | — |
case-20 | pass→pass | 25,745 | 25,089 | -3% | 1 | 1 | 0% | 4,599 | 4,717 | +3% | 0 | 0 | — |
case-21 | pass→fail | 24,237 | 30,075 | +24% | 1 | 1 | 0% | 3,566 | 5,163 | +45% | 0 | 0 | — |
case-22 | pass→pass | 10,084 | 5,225 | -48% | 1 | 1 | 0% | 1,941 | 1,491 | -23% | 0 | 0 | — |
case-23 | pass→pass | 13,283 | 11,452 | -14% | 1 | 1 | 0% | 2,672 | 2,808 | +5% | 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. 23 cases were attempted. The headline lift of 0 percentage points is the difference between those two pass rates over the 23 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.