skill-writer

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Skill Writer

Skill 编写工具

Use this as the single canonical workflow for skill creation and improvement. Primary success condition: maximize high-value input coverage before authoring so the resulting skill has minimal blind spots.

Load only the path(s) required for the task:

Task	Read
Set skill class and required dimensions	`references/mode-selection.md`
Apply writing constraints for depth vs concision	`references/design-principles.md`
Select structure pattern for this skill	`references/skill-patterns.md`
Select workflow orchestration pattern for process-heavy skills	`references/workflow-patterns.md`
Select output format pattern for deterministic quality	`references/output-patterns.md`
Choose workflow path and required outputs	`references/mode-selection.md`
Load representative synthesis examples by skill type	`references/examples/*.md`
Synthesize external/local sources with depth gates	`references/synthesis-path.md`
Author or update SKILL.md and supporting files	`references/authoring-path.md`
Optimize skill description and trigger precision	`references/description-optimization.md`
Iterate using positive/negative/fix examples	`references/iteration-path.md`
Evaluate behavior and compare baseline vs with-skill (opt-in quantitative)	`references/evaluation-path.md`
Register and validate skill changes	`references/registration-validation.md`

本工具是skill创建与优化的唯一标准工作流。核心成功条件：在编写前最大化覆盖高价值输入，确保最终生成的skill盲区最小。

仅加载任务所需的路径：

任务	读取路径
设置skill类别与必填维度	`references/mode-selection.md`
应用深度与简洁度的编写约束	`references/design-principles.md`
为当前skill选择结构模式	`references/skill-patterns.md`
为重流程类skill选择工作流编排模式	`references/workflow-patterns.md`
选择输出格式模式以保证确定性质量	`references/output-patterns.md`
选择工作流路径与必填输出	`references/mode-selection.md`
按skill类型加载代表性合成示例	`references/examples/*.md`
结合深度门槛合成外部/本地来源资源	`references/synthesis-path.md`
编写或更新SKILL.md及支撑文件	`references/authoring-path.md`
优化skill描述与触发精度	`references/description-optimization.md`
基于正/负/修复示例迭代	`references/iteration-path.md`
评估行为，对比基准表现与使用skill后的表现（可选量化）	`references/evaluation-path.md`
注册并验证skill变更	`references/registration-validation.md`

Step 1: Resolve target and path

步骤1：确定目标与路径

Resolve target skill path and intended operation (
```
create
```
,
```
update
```
,
```
synthesize
```
,
```
iterate
```
).
Read
```
references/mode-selection.md
```
and select the required path(s).

Classify the skill (

workflow-process

integration-documentation

security-review

skill-authoring

generic

Ask one direct question if class or depth requirements are ambiguous; otherwise state explicit assumptions.

确定目标skill路径与预期操作（
```
create
```
创建、
```
update
```
更新、
```
synthesize
```
合成、
```
iterate
```
迭代）。
阅读
```
references/mode-selection.md
```
并选择所需路径。
对skill进行分类（
```
workflow-process
```
工作流流程类、
```
integration-documentation
```
集成文档类、
```
security-review
```
安全审查类、
```
skill-authoring
```
skill编写类、
```
generic
```
通用类）。
如果类别或深度要求不明确，提出一个直接问题，否则明确说明假设条件。

Step 2: Run synthesis when needed

步骤2：必要时执行合成操作

Read

references/synthesis-path.md

Collect and score relevant sources with provenance.
Apply trust and safety rules when ingesting external content.
Produce source-backed decisions and coverage/gap status.
Load one or more profiles from
```
references/examples/*.md
```
when the skill is hybrid.
Enforce baseline source pack for skill-authoring workflows.
Enforce depth gates before moving to authoring.

阅读

references/synthesis-path.md

。

收集相关来源资源并标注来源进行评分。
摄入外部内容时遵守信任与安全规则。
输出有来源支撑的决策以及覆盖/缺口状态。
若skill为混合类型，从
```
references/examples/*.md
```
加载一个或多个参考模板。
为skill编写工作流强制执行基础资源包要求。
进入编写阶段前先完成深度门槛校验。

Step 3: Run iteration first when improving from outcomes/examples

步骤3：基于结果/示例优化时优先执行迭代

Read

references/iteration-path.md

first when selected path includes

iteration

(for example operation

iterate

Capture and anonymize examples with provenance.
Re-evaluate skill behavior against working and holdout slices.
Propose improvements from positive/negative/fix evidence.
Carry concrete behavior deltas into authoring.

Skip this step when selected path does not include

iteration

若所选路径包含

iteration

（例如

iterate

操作），请先阅读

references/iteration-path.md

。

捕获示例并匿名化处理，保留来源信息。
对照可用数据集与留存数据集重新评估skill表现。
基于正/负/修复证据提出优化方案。
将具体的行为变更带入编写阶段。

若所选路径不包含

iteration

则跳过此步骤。

Step 4: Author or update skill artifacts

步骤4：编写或更新skill产物

Read

references/authoring-path.md

Write or update
```
SKILL.md
```
in imperative voice with trigger-rich description.
Create focused reference files and scripts only when justified.

references/skill-patterns.md

references/workflow-patterns.md

, and

references/output-patterns.md

for structure and output determinism.

For authoring/generator skills, include transformed examples in references:
- happy-path
- secure/robust variant
- anti-pattern + corrected version

阅读

references/authoring-path.md

。

用祈使语气编写或更新
```
SKILL.md
```
，提供包含丰富触发词的描述。
仅在必要时创建聚焦的参考文件和脚本。

遵循

references/skill-patterns.md

、

references/workflow-patterns.md

和

references/output-patterns.md

的要求保证结构与输出确定性。

对于编写/生成类skill，在参考资料中包含转换后的示例：
- 正常路径示例
- 安全/鲁棒性变体
- 反模式 + 修正版本

Step 5: Optimize description quality

步骤5：优化描述质量

Read

references/description-optimization.md

Validate should-trigger and should-not-trigger query sets.
Reduce false positives and false negatives with targeted description edits.
Keep trigger language generic across Codex and Claude.

阅读

references/description-optimization.md

。

校验应当触发、不应当触发的查询集合。
通过针对性的描述编辑减少误报和漏报。
保持触发语言在Codex和Claude之间通用。

Step 6: Evaluate outcomes

步骤6：评估结果

Read

references/evaluation-path.md

Run a lightweight qualitative check by default (recommended).
For integration/documentation and skill-authoring skills, include the concise depth rubric from
```
references/evaluation-path.md
```
.
Run deeper eval playbook and quantitative baseline-vs-with-skill only when requested or risk warrants it.
Record outcomes and unresolved risks.

阅读

references/evaluation-path.md

。

默认运行轻量级定性检查（推荐）。
对于集成/文档类和skill编写类skill，引入
```
references/evaluation-path.md
```
中的简洁深度评分规则。
仅在收到要求或风险需要时运行更深层次的评估方案以及基准表现与使用skill后表现的量化对比。
记录结果与未解决的风险。

Step 7: Register and validate

步骤7：注册与验证

Read

references/registration-validation.md

Apply repository registration steps.
Run quick validation with strict depth gates.
Reject shallow outputs that fail depth gates or required artifact checks.

阅读

references/registration-validation.md

。

执行代码仓库注册步骤。
运行带有严格深度门槛的快速验证。
拒绝未通过深度门槛或必填产物检查的浅层次输出。

Output format

输出格式

Return:

```
Summary
```
```
Changes Made
```
```
Validation Results
```
```
Open Gaps
```

返回以下内容：

```
Summary
```
（概述）
```
Changes Made
```
（所做变更）
```
Validation Results
```
（验证结果）
```
Open Gaps
```
（未解决缺口）