重新验证

人工智能 社区

技能解读

当前原文的自动摘译,尚未人工审校。

第 1 步:结构验证

重新完整读取目标代理或技能配置文件。根据当前代码库验证每个记录的事实,包括当前会话更改:文件和目录路径存在并与文档匹配。存在引用的工具、参考文档和链接文件。命名类型、类、对话框和命名空间都是准确的。记录的约定仍然与现有代码匹配。

第二步:逻辑分析

分析机械路径检查之外的配置:职责与可用工具和参考文件一致。约定对于当前的代码库仍然有意义。各部分不会重叠、相互矛盾或重复陈旧的指导。涵盖了常用编辑区域。

流狐档案 运行与权限由流狐静态检查推断,不代表作者声明或实际测试结果。
领域
人工智能
作者声明的智能体
未找到明确声明;不据此推断已兼容或已测试
静态检查
88 / 100 · 启发式扫描,不代表运行安全
运行环境证据
未执行;当前没有隔离运行测试证据
作者 / 版本 / 许可证
@tomevault-io · 未声明许可证
上下文用量评级
低消耗
接入复杂程度
即装即用
是否需要外部接口密钥
不需要
兼容的系统
未声明(默认跨平台)
运行环境要求
未检出明确要求
检测到的文件与系统行为
  • 只读
  • 允许写入 / 修改
网络行为
仅限本地
安装命令数
无(仅作为资料)

需要注意: 未声明工具权限范围;实际权限由运行环境决定。

作者说明文件(原文) SKILL.md

# Agent And Skill Revalidation

Use this workflow to keep Codex configuration accurate and coherent with the actual codebase.

## Step 1: Structural Validation

Re-read the target agent or skill configuration file in full. Validate every documented fact against the current codebase, including current-session changes:

- File and directory paths exist and match the documentation.
- Referenced tools, reference docs, and linked files exist.
- Named types, classes, dialogs, and namespaces are accurate.
- Documented conventions still match existing code.

Apply objective structural fixes automatically. These corrections do not need approval.

## Step 2: Logical Analysis

Analyze the configuration beyond mechanical path checks:

- Responsibilities align with the available tools and reference files.
- Conventions still make sense for the current codebase.
- Sections do not overlap, contradict each other, or duplicate stale guidance.
- Commonly edited areas are covered.
- Self-maintenance rules catch likely side effects.
- Reference files are structured well enough to scale.

Present logical findings one at a time, ordered by importance. For each finding:

1. Describe the issue and concrete suggestion.
2. Ask the user to approve, reject, or modify the suggestion. Use a concise plain-text question when no structured question tool is available.
3. Apply approved changes immediately before moving to the next finding.

Do not batch logical findings. Do not apply logical-analysis changes without user approval.

---
> Source: [BitzArt-VS/UI-Tweaks](https://github.com/BitzArt-VS/UI-Tweaks) — distributed by [TomeVault](https://tomevault.io).
<!-- tomevault:4.0:skill_md:2026-05-23 -->

讨论

基于 GitHub Discussions。登录 GitHub 即可参与讨论、点赞、订阅更新。