一键导入
embroiderly-context
Use it to load and enforce domain glossary throughout the task.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Use it to load and enforce domain glossary throughout the task.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Use when the user wants to design or improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable or AI-navigable, or when another skill needs the deep-module vocabulary.
Use when writing, editing, or restructuring a guide, reference page, or any content under the `docs/` directory.
Use this skill when adding, editing, or reorganizing translations in the main app.
Make changes to the UI Kit. Use this skill when you need to create new or update existing UI components, their themes, stories, or tests.
Verify code quality after making changes in this project. Use this skill whenever you've finished implementing a feature, fix, or refactor and need to run the verification pipeline.
Use when the user wants to stress-test a plan before building, or uses any 'grill' trigger phrases.
| name | embroiderly-context |
| description | Use it to load and enforce domain glossary throughout the task. |
| allowed-tools | Read Write Edit |
| user-invocable | false |
CONTEXT_MAP.md at the repo root. If it doesn't exist yet, proceed without one and create it lazily when the first term is resolved.CONTEXT.md file(s) --- do not load all contexts upfront.Use only the canonical terms from the loaded glossary in all code, variable names, comments, documentation, and output.
When the user uses a term that conflicts with the glossary, surface it before proceeding: "Your glossary defines X as Y --- did you mean Z?"
When the user uses a vague or overloaded term not yet in the glossary, propose a precise canonical term rather than silently adopting their phrasing.
When a term is resolved --- whether clarified from vagueness or introduced for the first time --- update the relevant CONTEXT.md immediately.
Do not batch updates.
Create CONTEXT.md and CONTEXT_MAP.md lazily --- only when the first term is resolved, not speculatively.
Never create empty placeholder files.
CONTEXT.md template# {Context Name}
{One or two sentences: what this context is and why it exists}.
## Language
**{Term}**:
{One or two sentences: what it IS, not what it does}.
_Avoid_: {synonym}, {synonym}.
Rules:
_Avoid_._Avoid_.CONTEXT_MAP.md template# Context Map
## Contexts
- [{Name}](./{path}/CONTEXT.md) --- {one sentence: what this context owns}.
## Relationships
- **{Context A} -> {Context B}**: {how they relate --- events emitted/consumed, shared types, HTTP calls, etc.}.