write-a-skill
创建结构正确、支持 progressive disclosure 并带 bundled resources 的新 agent skills。Use when user wants to create, write, or build a new skill.
Skill by pop1414
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/pop1414-write-a-skill
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: writing, agent-building
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/write-a-skill
What's inside
- Skill name
write-a-skill- Description
- 创建结构正确、支持 progressive disclosure 并带 bundled resources 的新 agent skills。Use when user wants to create, write, or build a new skill.
- License
- MIT
- Instructions
- 114 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Writing Skills
## Process
1. **Gather requirements** — 询问用户:
- skill 覆盖什么 task/domain?
- 应处理哪些具体 use cases?
- 需要 executable scripts,还是只需要 instructions?
- 是否有 reference materials 要包含?
2. **Draft the skill** — 创建:
- 带 concise instructions 的 SKILL.md
- 如果内容超过 500 行,创建 additional reference files
- 如果需要 deterministic operations,创建 utility scripts
3. **Review with user** — 展示 draft 并询问:
- 是否覆盖你的 use cases?
- 是否缺失或不清楚?
- 是否有 section 应更详细或更简短?
## Skill Structure
```
skill-name/
├── SKILL.md # Main instructions (required)
├── REFERENCE.md # Detailed docs (if needed)
├── EXAMPLES.md # Usage examples (if needed)
└── scripts/ # Utility scripts (if needed)
└── helper.js
```
## SKILL.md Template
```md
---
name: skill-name
description: Brief description of capability. Use when [specific triggers].
---
# Skill Name
## Quick start
[Minimal working example]
## Workflows
[Step-by-step processes with checklists for complex tasks]
## Advanced features
[Link to separate files: See [REFERENCE.md](REFERENCE.md)]
```
## Description Requirements
description 是 agent 决定是否加载 skill 时**唯一看到的内容**。它会和其他 installed skills 一起出现在 system prompt 中。agent 会读取这些 descriptions,并根据用户请求选择相关 skill。
**Goal**:给 agent 足够信息,让它知道:
1. 这个 skill 提供什么 capability
2. 何时/为什么触发它(specific keywords、contexts、file types)
**Format**:
- 最多 1024 chars
- 使用 third person
- 第一句说明它做什么
- 第二句:"Use when [specific triggers]"
**Good example**:
```
Extract text and tables from PDF files, fill forms, merge documents. Use when working with PDF files or when user mentions PDFs, forms, or document extraction.
```
**Bad example**:
```
Helps with documents.
```
坏例子无法让 agent 区分它和其他 document skills。
## When to Add Scripts
以下情况添加 utility scripts:
- Operation 是 deterministic(validation、formatting)
- 同一段 code 会被反复生成
- Errors 需要明确 handling
相比 generated code,scripts 节省 tokens 并提升 reliability。
## When to Split Files
以下情况拆分为独立文件:
- SKILL.md 超过 100 行
- 内容有不同 domains(finance vs sales schemas)
- Advanced features 很少需要
## Review Checklist
draft 完成后验证:
- [ ] Description 包含 triggers("Use when...")
- [ ] SKILL.md 低于 100 行
- [ ] 没有 time-sensitive info
- [ ] Terminology 一致
- [ ] 包含 concrete examples
- [ ] References 只深入一层From pop1414/ai-native-doc-system, MIT.
More from pop1414/ai-native-doc-system
- 0.479score
git-guardrails-claude-code
pop1414
设置 Claude Code hooks,在危险 git commands(push, reset --hard, clean, branch -D 等)执行前阻止它们。Use when user wants to prevent destructive git operations, add git safety…
SkillCFreshness: CClaude Code +9 - 0.479score
diagnose
pop1414
面向棘手 bug、性能回退和 BMad story execution failures 的纪律化 diagnosis loop。Reproduce → minimise → hypothesise → instrument → fix → regression-test.
SkillCFreshness: CClaude Code +9 - 0.474score
caveman
pop1414
超压缩 communication mode。Cuts token usage ~75% by dropping filler, articles, and pleasantries while keeping full technical accuracy.
SkillCFreshness: CClaude Code +9 - 0.474score
to-issues
pop1414
将正式文档或 BMad 执行发现转成本地 issues。Use when user wants to convert a spec or plan into implementation issues, or create BMad control-panel issues for blockers, HITL…
SkillCFreshness: CClaude Code +9 - 0.474score
to-plan
pop1414
Converts a feature PRD, spec, BMad epics/stories, release goal, or conversation context into an execution plan at `docs/plans/<plan>.md` or…
SkillCFreshness: CClaude Code +9 - 0.474score
to-prd
pop1414
将当前 conversation context 或 BMad PRD 输出转成 `docs/features/<feature>/PRD.md`,并接入 AI-native 文档链。Use when user wants to create a PRD, write product requirements…
SkillCFreshness: CClaude Code +9 - 0.474score
to-spec
pop1414
Converts a feature PRD, plan, BMad architecture/epics/stories output, or conversation context into `docs/features/<feature>/SPEC.md` with functional behavior…
SkillCFreshness: CClaude Code +9 - 0.474score
grill-with-docs
pop1414
Grilling session that challenges a plan against the existing domain model, feature docs, and documented decisions, then updates CONTEXT.md and ADRs as…
SkillCFreshness: CClaude Code +9 - 0.474score
improve-codebase-architecture
pop1414
根据 CONTEXT.md 中的 domain language 和 docs/adr/ 中的 decisions,寻找 codebase 的 deepening opportunities。Use when the user wants to improve architecture, find…
SkillCFreshness: CClaude Code +9 - 0.474score
migrate-to-shoehorn
pop1414
将 test files 从 `as` type assertions 迁移到 @total-typescript/shoehorn。Use when user mentions shoehorn, wants to replace `as` in tests, or needs partial test data.
SkillCFreshness: CClaude Code +9 - 0.474score
scaffold-exercises
pop1414
创建包含 sections、problems、solutions 和 explainers 的 exercise directory structures,并确保通过 linting。Use when user wants to scaffold exercises, create exercise stubs…
SkillCFreshness: CClaude Code +9 - 0.474score
setup-ai-document-system
pop1414
Sets up repo-local adapters and `docs/agents/` configuration so engineering skills share one issue tracker, triage vocabulary, domain layout, AI-native…
SkillCFreshness: CClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Jun 1, 2026 (5 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/pop1414/ai-native-doc-system
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from pop1414
- 0.479score
git-guardrails-claude-code
pop1414
设置 Claude Code hooks,在危险 git commands(push, reset --hard, clean, branch -D 等)执行前阻止它们。Use when user wants to prevent destructive git operations, add git safety…
SkillCFreshness: CClaude Code +9 - 0.479score
diagnose
pop1414
面向棘手 bug、性能回退和 BMad story execution failures 的纪律化 diagnosis loop。Reproduce → minimise → hypothesise → instrument → fix → regression-test.
SkillCFreshness: CClaude Code +9 - 0.474score
improve-codebase-architecture
pop1414
根据 CONTEXT.md 中的 domain language 和 docs/adr/ 中的 decisions,寻找 codebase 的 deepening opportunities。Use when the user wants to improve architecture, find…
SkillCFreshness: CClaude Code +9 - 0.474score
grill-with-docs
pop1414
Grilling session that challenges a plan against the existing domain model, feature docs, and documented decisions, then updates CONTEXT.md and ADRs as…
SkillCFreshness: CClaude Code +9 - 0.474score
caveman
pop1414
超压缩 communication mode。Cuts token usage ~75% by dropping filler, articles, and pleasantries while keeping full technical accuracy.
SkillCFreshness: CClaude Code +9 - 0.474score
migrate-to-shoehorn
pop1414
将 test files 从 `as` type assertions 迁移到 @total-typescript/shoehorn。Use when user mentions shoehorn, wants to replace `as` in tests, or needs partial test data.
SkillCFreshness: CClaude Code +9
Cite this page
write-a-skill. Agent Mart. https://agentmart.dev/skills/pop1414-write-a-skill (last updated 2026-10-24)