sgc-skill-distiller SKILL.md skillA

sgc-skill-distiller from lovstudio/skills is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned
mdr add lovstudio/skills/sgc-skill-distiller@v0.4.0

Pins v0.4.0 in mdr.lock. Pin the fingerprint instead.

Audit grade
A, 17 of 17 checks passed
Kind
SKILL.md skill
Repository
lovstudio/skills
Path
skills/skill-distiller/SKILL.md
Versions recorded
4
Latest version
v0.4.0, committed Sep 7, 2026
First version recorded
Aug 10, 2026
Size
5,325 bytes
Repository stars
72
License
MIT

Use when 用户要将项目经验、故障复盘和已验证流程蒸馏为可创建的 Agent Skill 蓝图,明确用户结果、触发边界、私有信息边界与验收方式;触发词包括“把经验蒸馏成 skill”与 “distill experience into a skill”。.

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# Skill 炼金师 · Skill Alchemist

这不是泛泛寻找机会,而是把已经发生过的工作蒸馏成稳定、可移植且可验收的能力契约。最终产物是 **Skill 蓝图**:它足以交给 `sgc-skill-creator` 落地,却不泄漏人名、项目代号、私有路径、密钥或聊天背景。

## Triggers

### Activate when

- 用户说“把这段经验蒸馏成 Skill”“把 Yoda 的做法沉淀下来”或“把这次复盘做成可复用能力”。
- 用户提供项目历史、事故复盘、连续需求或真实交付,想从中抽出一项可复用 Skill。
- Use this when the user says “distill experience into a skill”, “turn this workflow into a reusable skill”, or “skill distillation”.

### Do not activate when

- 用户已明确要实现某一个 Skill:交给 `sgc-skill-creator`。
- 用户要优化、修订或升级已有 Skill:交给 `sgc-skill-optimizer`。
- 用户只要保存一条项目约定到系统提示词或项目说明:交给 `sgc-distill-to-system`。

## Workflow (MANDATORY)

### Step 0: 选择待蒸馏的经验单元

以一个完整的“问题 → 判断 → 行动 → 验证”闭环为单位。优先选择经过真实运行、发布、安装或用户验收的经验;单次灵感、纯技术兴趣或只含内部背景的内容暂不提升为 Skill。

把人名、客户信息、项目代号、私有路径、访问凭据和临时聊天细节留在证据层,默认不进入蓝图的名称、描述和示例。

### Step 1: 收集原始证据(辅助,不是主产物)

针对仓库或材料创建可回看的证据摘要:

```bash
python3 "$SKILL_DIR/scripts/collect-source-evidence.py" PROJECT_PATH \
  --output skill-distillation-evidence.md
```

Git 历史只是线索。补充用户提供的复盘、验收记录、失败日志与行为变化;没有证据支撑的判断要标注为假设。

### Step 2: 蒸馏不变量

使用 [蒸馏透镜](references/distillation-lens.md),从原始材料中抽出:

1. **用户真正要达成的结果**,而不是内部实现动作;
2. **稳定的判断顺序**:哪些步骤总是成立,哪些是可配置的变体;
3. **失败恢复与验收**:如何知道“真的成功”,而不是只看到提示、编译或静态检查;
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint
mdr add lovstudio/skills/sgc-skill-distiller@sha256:074bf382bbb4ea6d

The line at the top of this page pins the label v0.4.0, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_6gwmvdqdid3nf42b.svg)](https://markdownregistry.com/a/art_6gwmvdqdid3nf42b)

2 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 4 versions of this file, between Aug 10, 2026 and Sep 7, 2026, about one every 9 days.

versioncommittedcommitsizeaudit
v0.4.0 latest2026-09-07 a09848e 5,325 BA view · diff
v0.4.02026-08-23 00fafa8 5,348 BA view · diff
v0.3.12026-08-11 e054f41 4,626 BA view · diff
v0.3.12026-08-10 8d07ad2 4,626 BA view

Audit of the latest version

A  17 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (5325 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

lovstudio/skills · 72 stars · license MIT · pushed 2026-11-06 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_6gwmvdqdid3nf42b
GET https://markdownregistry.com/api/v1/resolve?ref=lovstudio/skills/sgc-skill-distiller
GET https://markdownregistry.com/api/v1/blob/074bf382bbb4ea6d3eebc5935fdcbcceaa87e6914df9dd3e34fd432192150860

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.

More from lovstudio/skills

CLAUDE.md CLAUDE.md file
lovstudio/skills · CLAUDE.md
git:20261031.adbb937 · audit A · 72 stars
lov-academic-translator SKILL.md skill
lovstudio/skills · skills/academic-translator/SKILL.md · 将英文论文、arXiv、期刊/会议文章及技术 PDF 翻译为中文 PDF 或可发布的中文 Markdown。用户提出“翻译英文 PDF”“翻译论文”“论文超级翻译官”“保留图片、公式和原版式”“原文译文对照”“页数对应”或“PDF
v0.2.0 · audit A · 72 stars
lov-add-task SKILL.md skill
lovstudio/skills · skills/add-task/SKILL.md · 在当前任务清单中按优先级和位置追加事项并保留已有状态。支持明确输入与结果回读。Use to add an item to the current
v0.1.2 · audit A · 72 stars
lov-add-tool SKILL.md skill
lovstudio/skills · skills/add-tool/SKILL.md · 为现有网站新增可用的在线工具,并接入多语言、工具目录、站点地图和必要的成本提示。支持明确输入与结果回读。Use to add an online
v0.1.2 · audit A · 72 stars
lov-agent-instructions SKILL.md skill
lovstudio/skills · skills/agent-instructions/SKILL.md · 诊断或完善项目 Agent 指令文件,适配不同宿主并保留真实工程约定。支持明确输入与结果回读。Use to review or improve
v0.1.2 · audit A · 72 stars
lov-algo-viz-creator SKILL.md skill
lovstudio/skills · skills/algo-viz-creator/SKILL.md · 为算法、数据结构或信息流程生成「逐步演示」的可交互可视化。输入:算法名或 概念描述(如“把 KMP 做成可视化”)。输出:自包含单文件 HTML,支持播放/暂停/ 逐帧/变速与键盘演示,每步附中文解说与伪代码高亮。 触发语:做成可视化 /…
v0.1.1 · audit A · 72 stars
lov-align-wechat-uiux SKILL.md skill
lovstudio/skills · skills/align-wechat-uiux/SKILL.md · 系统诊断并修复产品与微信官方在消息、会话、联系人、朋友圈及交互上的语义和 UIUX 偏差;适用于“对齐微信官方”“这条微信消息显示不对”、"fix WeChat UI parity" 或 "audit WeChat behavior"。
v0.2.1 · audit A · 72 stars
lov-wechat-evidence SKILL.md skill
lovstudio/skills · skills/align-wechat-uiux/skills/wechat-evidence/SKILL.md · 为微信体验差异建立可追溯证据集,关联官方表现、当前产品、原始记录、代码路径与版本环境;适用于“先确认微信官方怎么表现”、"review the WeChat reference" 或 parity evidence audit。
v0.1.1 · audit A · 72 stars
lov-wechat-parity-validation SKILL.md skill
lovstudio/skills · skills/align-wechat-uiux/skills/wechat-parity-validation/SKILL.md · 验证微信体验修复是否同时覆盖解析、共享表现、视觉交互、下游消费者与真实运行时;适用于“验收微信对齐”“还有没有漏改”、"review the parity fix" 或 validate WeChat behavior。
v0.1.1 · audit A · 72 stars
lov-wechat-semantics SKILL.md skill
lovstudio/skills · skills/align-wechat-uiux/skills/wechat-semantics/SKILL.md · 将微信数据库类型、嵌套 XML/JSON、身份与状态归一为稳定产品语义,定位最早失真层;适用于“这是什么微信消息”“修复消息分类”、"model the WeChat message" 或 review parser semantics。
v0.1.1 · audit A · 72 stars
lov-wechat-ui-parity SKILL.md skill
lovstudio/skills · skills/align-wechat-uiux/skills/wechat-ui-parity/SKILL.md · 在既有设计系统内实现微信官方的消息层级、组件语义、交互状态与视觉密度;适用于“按微信官方改这个组件”“还原微信聊天体验”、"build the WeChat UI behavior" 或 fix official UI parity。
v0.1.1 · audit A · 72 stars
lov-anti-wechat-ai-check SKILL.md skill
lovstudio/skills · skills/anti-wechat-ai-check/SKILL.md · 兼容旧命令的中文表层模式扫描器,只定位套话、句段均匀和过渡词等可观察特征, 不判断作者身份或平台风险。新任务请使用 lov-human-writing。Use only for legacy “anti-ai-check”…
v1.1.2 · audit A · 72 stars

Every file in lovstudio/skills

Browse by kind, by grade A, or by owner.