Claude Code
Anthropic's official terminal coding agent — reads and edits your whole repo, runs commands, and handles multi-step dev work. This section collects CLAUDE.md, custom slash commands, hooks, subagents, MCP, Agent Skills, and TDD/debugging workflows — original techniques plus curated open-source (official + high-star MIT).
PROMPTS
Claude Code-specific tested prompts. Click through for full version, variables, and usage context. Prompt bodies are authored in Traditional Chinese.
Claude Code 安全稽核員 Subagent:OWASP Top 10 + 權限/認證漏洞掃描
把一個專注的「安全稽核員」子代理裝進 Claude Code,在功能開發當下就用 OWASP Top 10 的角度逐檔審查程式碼與架構,按嚴重度分類漏洞並附上可直接套用的修補建議。
Claude Code 後端架構師 Subagent:API 契約先行 + 微服務邊界設計
把一個資深「後端架構師」子代理裝進 Claude Code,用契約先行(contract-first)與領域驅動設計來切服務邊界、設計 REST/GraphQL/gRPC API,並從第一天就把韌性與可觀測性納入架構。