全局重度代码评审

SkillFiles & storage

Performs a read-only, heavyweight code review of the entire VLink repository or a specified scope. Plans coverage and subagents by feature, checks each file for correctness, concurrency, performance, and standards, and delivers evidence-based conclusions in Chinese after adversarial re-review. Use w

Available today. Use it from your connected AI after setup.

Connect ahel once, and every AI you use reads what you have installed.

Then ask your AI: use the 全局重度代码评审 skill

What this skill tells your AI

The instructions your AI receives, as published by thun-res/vlink in .agents/skills/deep-review/SKILL.md and read by ahel’s review.

本 skill 只评审,不修改源码。用户指定范围时审计对应子集;未指定时覆盖 全部第一方受管文件;只有明确要求落盘时才写 .agents/cache-report/

1. 功能路由

只加载当前阶段需要的分册:

功能必读分册
建立文件全集、模块边界和排除项SCOPE-PLANNING.md
逐文件检查并判定严重级REVIEW-DIMENSIONS.md
并行派发与对抗复核AGENT-ORCHESTRATION.md
汇总证据、覆盖率和报告REPORTING.md

2. 执行顺序

  1. 固化审计范围和文件清单,再按职责与依赖切分任务。
  2. 每个任务逐文件审查,返回覆盖清单和带行号发现。
  3. 汇总后回读源码核实所有 P0/P1,对重要结论做反驳式复核。
  4. 跨模块去重,列出未覆盖项;明确要求落盘时写入报告。

不得以抽样代替承诺的全量覆盖,不得把“未发现”写成“已证明没有问题”。

Signals

GitHub stars
115
Forks
10
Last commit
Sep 2026
Advanced
Catalog kind
skill
Gateway key
deep-review-thun-res
Source
github.com/thun-res/vlink