一键导入
api-contract-writer
Design, review, evolve, and document API contracts with stable schemas, examples, error models, versioning, and compatibility rules.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Design, review, evolve, and document API contracts with stable schemas, examples, error models, versioning, and compatibility rules.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | api-contract-writer |
| description | Design, review, evolve, and document API contracts with stable schemas, examples, error models, versioning, and compatibility rules. |
Use this preset for contract-first API work. The job is not just to describe a payload. The job is to define an interface that can be implemented, integrated, reviewed, and evolved safely.
Choose one primary mode per request:
| Mode | Main deliverable |
|---|---|
design | A clean contract draft with schemas, semantics, and examples |
review | Findings first: ambiguity, breakage, missing semantics, mismatches |
evolve | Change plan with compatibility analysis and migration notes |
document | Usage guidance, examples, integration notes, field semantics |
test | Contract test cases, validation rules, edge cases, breaking-change checks |
Do not force every response into a full spec. Match the user's actual ask.
When drafting a contract:
Prioritize:
Always call these out explicitly:
Distinguish clearly between additive, conditionally compatible, and breaking changes.
When the user asks for a substantial contract artifact, aim to include:
If the user asks for one narrow piece, give that piece without padding the answer with boilerplate.
Before finishing, verify:
专业的代码重构师,从代码分析开始,到代码架构的重编排。将代码的逻辑按图论解析为图,在逻辑层优化该图运行,并结束这一切。
专业的代码审查技能,支持多语言代码的全面性能分析和质量评估。包含详细的磁盘IO、网络IO、死锁风险分析,统一的接口规范和注释标准检查。用于代码Review、性能优化建议、代码质量评估等场景。
Solve, review, optimize, generate, and test algorithm problems with constraint-driven reasoning. Use when the user asks for algorithm design, competitive programming help, LeetCode-style practice, correctness review, complexity reduction, or adversarial test generation.