Vibe Coding、Cursor、Claude Code、Codex、Copilot
This collection features articles and community discussions in the AI 编程 space, focusing on real-world engineering scenarios, tool selection, practical paths, and risk boundaries.
The page is continuously updated and serves as a learning roadmap, technology selection reference, and technical review entry point.
Direct fixes to generated artifacts usually disappear during the next codegen run. This guide uses ownership manifests, deterministic regeneration, extension layers, and atomic publishing to define who can change sources or templates and how to roll back an entire artifact set safely.
Search results do not prove that they belong to the current commit. This article uses commit watermarks, change logs, deletion and rename handling, dirty-worktree overlays, and atomic index publication to prevent coding agents from reading stale symbols.
Manual code review is a team bottleneck—PRs pile up, reviews are hasty, style arguments drain energy. AI review does not replace people; it filters the mechanical, easy-to-miss issues first so people can focus on architecture and business. Covers what AI review can do, what it can't replace, PR integration, and noise-control practices.
Git worktrees isolate indexes and working directories, but do not automatically resolve parallel ownership of files, generated targets, and shared contracts. This article uses versioned claims, monotonic generations, diff admission, and ref CAS to create a reclaimable integration protocol that late agents cannot overwrite.
Directory proximity does not imply a dependency. This article organizes build, symbol, configuration, and runtime edges into a versioned repository graph so coding Agents can compute the impact closure, rollout order, and rollback boundaries before making changes.
AI writing assistants are easy to imagine but easy to get wrong — most become 'one-click filler generators.' From requirements to launch, this article explains how to design a genuinely useful assistant: assist not ghostwrite, structured streaming, a human-AI editing loop, and trading 'one-click generation' for 'better the more you write.'
Rate limiting for AI apps is more complex than for web apps: there are two cost dimensions (requests and tokens), plus abuse, automated traffic, and single-user spikes. This article covers three layers of limits (user, IP, and global), token quota accounting, graceful degradation, and a quota architecture resilient to malicious usage.
Prompts contain core AI application logic, yet teams often edit them as arbitrary "configuration strings," causing production failures, missing rollback history, and unusable A/B tests. This article explains how to manage prompts like code through versioning, CI evaluations, gradual rollout, and rollback.
How do you know whether a new prompt or model is actually better? AI A/B testing must account for nondeterministic outputs, subjective quality, and large sample requirements. This guide covers metrics, offline evaluations, online experiments, and how to replace intuition with measurable product outcomes.
Software engineering has TDD; AI engineering needs EDD, or eval-driven development. Build an eval set first to define "good," run it on every change, and block regressions. This article covers the workflow, its nondeterministic and probabilistic differences from TDD, implementation, and how EDD turns intuition-led AI tuning into an engineering feedback loop.
Copilot, Cursor, and Claude Code are great—but can you build your own code generator? This article breaks down the engineering core: context construction, generation strategy (completion/edit/agent), quality guardrails (syntax/test/diff), and how an internal code assistant differs from a general-purpose AI tool: scenario focus plus codebase awareness.
Test debt hurts every team, and AI test generation can ease it—but also produce meaningless tests and fake coverage. Covers the right way: which tests suit AI, the generate→validate→accept workflow, avoiding the coverage trap, tool selection (Claude Code/Codium/in-house), and the parts AI can't replace.