Yao-meta-skill yao-meta-skill

Create, refactor, evaluate, and package agent skills from workflows, prompts, transcripts, docs, or notes. Use when asked to create a skill, turn a repeated process into a reusable skill, improve an existing skill, add evals, or package a skill for team reuse.

install

source · Clone the upstream repo

git clone https://github.com/yaojingang/yao-meta-skill

Claude Code · Install into ~/.claude/skills/

git clone --depth=1 https://github.com/yaojingang/yao-meta-skill ~/.claude/skills/yaojingang-yao-meta-skill-yao-meta-skill

manifest: SKILL.md

source content

Yao Meta Skill

Build reusable skill packages, not long prompts.

Router Rules

Route by frontmatter
```
description
```
first.
Keep
```
SKILL.md
```
to routing plus a minimal execution skeleton.
Put long guidance in
```
references/
```
, deterministic logic in
```
scripts/
```
, and evidence in
```
reports/
```
.
Use the lightest process that still makes the skill reliable.

Modes

```
Scaffold
```
: exploratory or personal use.
```
Production
```
: team reuse with focused gates.
```
Library
```
: shared infrastructure or meta skill.

Mode rules: Operating Modes, QA Ladder, Resource Boundary Spec, Method.

Compact Workflow

Decide whether the request should become a skill, then choose the lightest fit.
Run a short intent dialogue to capture the real job, outputs, exclusions, constraints, and standards.
Run a reference scan: external benchmarks first, user references second, local fit checks third.
Write the
```
description
```
early and test route quality before expanding the package.
Add only the folders and gates that earn their keep.
After the first package exists, surface the top three next iteration directions.

Core playbooks: Method, Intent Dialogue, Reference Scan, Archetypes, Gate Selection, Iteration Philosophy, Non-Skill Decision Tree.

First-Turn Style

When the skill first activates:

open warmly, like a thoughtful teacher or design partner
start from the user's work and desired outcome before asking for structure
ask only
```
2-3
```
high-leverage questions unless the user already gave enough detail
let the user answer naturally first; offer a tiny scaffold only as an optional shortcut
do not default to cold field lists such as
```
Name / Capability / Inputs / Outputs
```

Chinese conversations should sound soft and companion-like rather than procedural.

For concrete opening patterns, see Intent Dialogue.

Output Contract

Unless the user asks otherwise, produce:

a working skill directory
a
```
SKILL.md
```
aligned
```
agents/interface.yaml
```

optional

references/

scripts/

evals/

reports/

, and

manifest.json

only when justified

a short summary of boundary, exclusions, references, gates, and next steps

Reference Map

Primary references: Method, Reference Scan, Intent Dialogue, Governance, Resource Boundaries.