- coding skill 发现、解析与注入边界
Scanned 8/30/2026
Install to Claude Code
npx -y skills add zhnt/loushang --skill component-interfaces --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Component Interfaces?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/zhnt-component-interfaces)More formats (shields.io, HTML) on the badges page.
# `skill`
## Role
- coding skill 发现、解析与注入边界
## Owns
- `SkillLoader`
- skill 元数据解析
- skill prompt 片段收集
- skill 可见性与启用策略
## Depends On
- `loader`
- `utils`
## Commands
- `discover_skills(...)`
- `load_skill(...)`
- `reload_skills(...)`
- `enable_skill(...)`
- `disable_skill(...)`
## Queries
- `get_skill(...)`
- `list_skills()`
- `list_enabled_skills()`
## Events
- 当前无稳定事件面
## Key Data
- `SkillDescriptor`
- `ResourceBundle`
- `SKILL.md` frontmatter:
- `name`
- `description`
- `disable-model-invocation`
## Out Of Scope
- method 选择策略
- extension hook 执行
- prompt 最终组装
## Reference Implementation Alignment
- 语义上吸收 `reference CLI` 的 customization / resource discovery 经验
- 在 Python 设计里显式保留 `SkillLoader`,让 skill 解析边界更清楚
- 但它更适合作为 resource loader 体系内的显式子边界,而不是另一套并列资源中心
- `description` 用于 `/skill:name` command 描述与 system prompt 中的 available skills 摘要。
- `disable-model-invocation: true` 让 skill 仍可显式 `/skill:name` 调用,但不会进入模型可自动发现的 skill 摘要。
- skill discovery 递归扫描 `skills/**/SKILL.md`;一旦某目录本身包含 `SKILL.md`,该目录即为 skill root,不再继续向下递归。
- discovery 跳过隐藏目录和 `node_modules`,并读取 `.gitignore` / `.ignore` / `.fdignore` 中的常用目录/路径/glob 忽略规则,避免把依赖、隐藏工作区或生成目录误注册为可用 skill。
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!