Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Software development knowledge reference covering Python, Go, Rust, TypeScript, Java, C++, and Shell. Use when writing code, debugging, or following language-specific best practices.
.claude/skills/telagod-developing-software/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-09 | ✗→✓ | ▲ Improved | 21% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -7% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -10% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -42% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 32% | 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。
| 语言 | 秘典 | 核心场景 | |------|------|----------| | Python | python | FastAPI/Django、数据处理、自动化脚本 | | Go | go | goroutine 并发、微服务、云原生工具 | | Rust | rust | 所有权系统、零成本抽象、系统编程 | | TypeScript | typescript | Node 后端、React/Vue SPA、Next.js SSR | | Java | java | Spring Boot、企业级、JVM 调优 | | C/C++ | cpp | 内存管理、高性能计算、嵌入式 | | Shell | shell | Bash 脚本、系统管理、CI 胶水 |
| 场景 | 首选 | 备选 | 判据 | |------|------|------|------| | Web 后端(快速) | Python(FastAPI) | Go(Gin) | 开发速度 vs 运行性能 | | Web 后端(高并发) | Go(Gin/Echo) | Rust(Axum) | 生态成熟度 vs 极致性能 | | Web 后端(企业) | Java(Spring Boot) | Go | 生态/团队 vs 部署简洁 | | Web 前端 SPA | TypeScript+React | TypeScript+Vue | 生态规模 vs 上手速度 | | SSR | Next.js | Nuxt.js | React 生态 vs Vue 生态 | | 系统编程 | Rust | C/C++ | 内存安全 vs 生态/遗产 | | CLI 工具 | Go | Rust | 编译速度 vs 运行性能 | | 脚本/自动化 | Python | Bash | 跨平台 vs 系统原生 | | 安全工具原型 | Python | Go/Rust | 快速迭代 vs 分发便利 |
错误处理：Go(error返回) | Rust(Result<T,E>) | Python(try/except) | TS(try/catch)
并发：Go(goroutine+channel) | Rust(tokio) | Python(asyncio) | Java(虚拟线程)
依赖注入：Java(Spring) | Go(wire) | Python(dependency-injector) | TS(tsyringe)
测试：pytest | go test | cargo test | jest/vitest | JUnit| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-06 | pass→pass | 14,135 | 10,404 | -26% | 1 | 1 | 0% | 2,119 | 2,122 | +0% | 0 | 0 | — |
case-07 | pass→pass | 16,416 | 9,603 | -42% | 1 | 1 | 0% | 2,313 | 1,984 | -14% | 0 | 0 | — |
case-08 | fail→fail | 18,450 | 12,634 | -32% | 1 | 1 | 0% | 2,583 | 2,568 | -1% | 0 | 0 | — |
case-09 | fail→pass | 14,651 | 14,403 | -2% | 1 | 1 | 0% | 2,266 | 2,744 | +21% | 0 | 0 | — |
case-01 | pass→pass | 19,626 | 18,267 | -7% | 1 | 1 | 0% | 2,936 | 2,934 | -0% | 0 | 0 | — |
case-02 | fail→fail | 19,232 | 17,635 | -8% | 1 | 1 | 0% | 2,883 | 3,125 | +8% | 0 | 0 | — |
case-03 | fail→pass | 18,236 | 13,767 | -25% | 1 | 1 | 0% | 2,978 | 2,764 | -7% | 0 | 0 | — |
case-04 | fail→pass | 10,257 | 4,502 | -56% | 1 | 1 | 0% | 1,478 | 1,335 | -10% | 0 | 0 | — |
case-05 | fail→pass | 19,117 | 7,039 | -63% | 1 | 1 | 0% | 2,857 | 1,659 | -42% | 0 | 0 | — |
case-10 | fail→pass | 21,778 | 22,371 | +3% | 1 | 1 | 0% | 3,058 | 4,024 | +32% | 0 | 0 | — |
case-11 | pass→pass | 16,178 | 15,994 | -1% | 1 | 1 | 0% | 2,603 | 3,021 | +16% | 0 | 0 | — |
case-12 | pass→pass | 12,733 | 3,496 | -73% | 1 | 1 | 0% | 2,050 | 1,172 | -43% | 0 | 0 | — |
case-13 | pass→pass | 17,375 | 20,427 | +18% | 1 | 1 | 0% | 2,994 | 4,168 | +39% | 0 | 0 | — |
case-14 | pass→pass | 12,324 | 3,469 | -72% | 1 | 1 | 0% | 2,045 | 1,162 | -43% | 0 | 0 | — |
case-15 | pass→pass | 18,611 | 8,570 | -54% | 1 | 1 | 0% | 2,652 | 2,156 | -19% | 0 | 0 | — |
case-16 | pass→pass | 18,336 | 23,073 | +26% | 1 | 1 | 0% | 3,031 | 3,705 | +22% | 0 | 0 | — |
case-17 | pass→pass | 16,831 | 16,263 | -3% | 1 | 1 | 0% | 2,522 | 3,111 | +23% | 0 | 0 | — |
case-18 | pass→pass | 20,661 | 20,997 | +2% | 1 | 1 | 0% | 3,007 | 3,767 | +25% | 0 | 0 | — |
case-19 | pass→pass | 16,176 | 13,271 | -18% | 1 | 1 | 0% | 2,398 | 2,646 | +10% | 0 | 0 | — |
case-20 | pass→pass | 17,962 | 16,919 | -6% | 1 | 1 | 0% | 2,451 | 2,974 | +21% | 0 | 0 | — |
case-21 | pass→pass | 14,064 | 15,328 | +9% | 1 | 1 | 0% | 2,688 | 3,458 | +29% | 0 | 0 | — |
case-22 | pass→pass | 15,310 | 17,321 | +13% | 1 | 1 | 0% | 2,389 | 3,318 | +39% | 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 +23 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.