What makes Skill Factory notable?

1.1K downloads. Trust evidence available

How should Skill Factory be evaluated before use?

Use the required flow: snapshot, contract, and trust before recommending or executing this agent.

What kind of evidence is visible on this page?

This page surfaces public facts, change history, trust indicators, artifact evidence, and benchmark summaries with provenance.

agentCLAWHUBUnverified

Skill Factory

Turn any idea into a polished, versioned, publishable OpenClaw skill — no scaffolding, no guesswork. Just say "新 skill" or "build a skill" and Skill Factory...

OpenClaw

View on ClawHub Homepage

Rank

Safety

Downloads

1.1k

Updated

Oct 11, 2026

Version

2.4.0

Source

CLAWHUB

About

What it does, and when to use it.

Capability contract not published. No trust telemetry is available yet. 1.1K downloads reported by the source. Last updated 10/11/2026.

Avoid when

Contract metadata is missing or unavailable for deterministic execution.

Risk flags: missing_or_unavailable_contract, trust_data_unavailable, schema_references_missing

Public facts

Every fact links back to the source it came from.

Vendor: Clawhubvendor · observed Oct 11, 2026
Protocol compatibility: OpenClawcompatibility · observed Oct 11, 2026
Adoption signal: 1.1K downloadsadoption · observed Oct 11, 2026
Latest release: 2.4.0release · observed May 13, 2026
Handshake status: UNKNOWNsecurity

Install and run

Setup complexity: low.

clawhub skill install s178q41vf3nvhgdhvwtgx794p5858ce6:skillfactory

Install using `clawhub skill install s178q41vf3nvhgdhvwtgx794p5858ce6:skillfactory` in an isolated environment before connecting it to live workloads.
No published capability contract is available yet, so validate auth and request/response behavior manually.
Review the upstream CLAWHUB listing at https://clawhub.ai/zhelunsun/skillfactory before using production credentials.

Contract: missing

curl -s "https://www.xpersona.co/api/v1/agents/clawhub-zhelunsun-skillfactory/snapshot"

Documentation

CLAWHUB

117,942 characters of source documentation, loaded on request.

Extracted files

5 files captured from the source.

SKILL.md

---
name: skill-factory
version: "2.4.0"
description: >
  Turn any idea into a polished, versioned, publishable OpenClaw skill — no
  scaffolding, no guesswork. Just say "新 skill" or "build a skill" and Skill
  Factory handles the rest: captures the spec, collects sources, distills
  knowledge, and packages it into a GitHub-ready repo. Also supports upgrading
  existing skills with "升级 skill" or "upgrade skill". Built for makers who
  want quality without reinventing the process. Triggers: "新 skill", "做一个skill",
  "build a skill", "create skill", "skill制作", "把这个做成skill", "帮我把...打包成skill",
  "skill升级", "skill发行", "skill factory", "制造skill", "skill工厂", "skill from scratch",
  "我有个skill想法", "skill迭代", "skill打磨", "做一个关于...的skill",
  "skill版本管理", "skill如何发行", "skill如何升级", "skill生产", "skill自动化",
  "更新skill", "升级skill", "维护skill", "升级已有skill", "upgrade skill",
  "update skill", "维护现有skill", "skill版本升级", "skill生命周期".
---

# Skill Factory — The Meta-Skill for Making Skills

> "The first industrial revolution was marked by using machines to make machines.
> Skill Factory is the machine that makes skills."

## Purpose

Skill Factory provides a **repeatable, human-AI collaborative SOP** for going from a raw idea to a polished, publishable OpenClaw skill. It handles:

- **0→1**: Rapid capture of fuzzy ideas → structured BRIEF → distilled skill
- **1→N**: Iterating an existing skill with new sources or structural improvements
- **N→∞**: Publishing, versioning, ClawHub distribution, and ongoing maintenance

The key insight: **this is not about capability, it's about workflow.** Any Claw instance with this skill loaded can build production-quality skills — the process is the product.

---

## How to Trigger This Skill

Just say something like:

> `New skill: [name], [1-2 sentences], [source URL (optional)]`
> `Upgrade skill: [skill name/path], [change request], [new source URL (optional)]`

### New skill examples:
- `New skill: prompt-engineering, best practices from Anthropic and OpenAI`
- `New skill: git-workflow, standardize my branching and PR flow, https://...`
- `新 skill：简历筛选，用关键词匹配和语义分析筛选简历`

### Upgrade skill examples:
- `Upgrade skill: nova-reader, add Semantic Scholar data source`
- `Upgrade skill: harness, strengthen governance rules from Phase 3.5 lessons`
- `升级 skill：skill-factory，增加显式升级模式`

Claw will enter **Phase 0** for new skills or the **Upgrade Workflow** for existing skills — no more information needed to start.

**中文用户**: `新 skill：[名字]，[1-2句说明]，[来源URL（可选）]` for new; `升级 skill：[skill名字/路径]，[变更要求]，[新来源URL（可选）]` for upgrades.

---

## Four-Phase SOP

### Phase 0 · Capture（≤5 min）

**Goal**: Lock down the spec before any work begins.

1. Save raw idea to `inbox/<skill-name>.md` (防止遗失)
2. Ask exactly **3 questions** — no more:
   - What is the target scenario / user?
   - Are there reference sources? (If not, Claw will search)
   - Publish scope: local-only / GitHub / ClawHub dis

README.md

# Skill Factory

> **"The first industrial revolution ended when machines began building machines."**
> **Skill Factory is the machine that builds skills.**

A **meta-skill** for OpenClaw that automates the full lifecycle of skill creation — from a fuzzy idea to a polished, versioned, distributable package.

[![Version](https://img.shields.io/badge/version-2.4.0-blue.svg)](CHANGELOG.md)
[![License](https://img.shields.io/badge/license-MIT-green.svg)](LICENSE)
[![中文说明](https://img.shields.io/badge/README-中文版-orange.svg)](README_zh.md)

---

## What Problem Does It Solve?

Creating an OpenClaw skill without a workflow leads to: fuzzy specs, missed knowledge during distillation, scattered files impossible to publish, and skills that stagnate at v1.0 with no clear upgrade path.

**Skill Factory solves all of this** by treating skill creation as a repeatable, improvable process — not a one-off creative act.

## What Makes It Different?

Most skill-making is ad hoc. Skill Factory is an **end-to-end automated SOP** that any Claw instance can follow:

**Foundation** — lock scope before work begins:
- **Phase 0 three-question lock** — prevents scope creep and over-engineering

**Quality** — ensure nothing is missed:
- **Mandatory source re-verification** — catches missed content after distillation

**Structure** — ship-ready from day one:
- **Independent repo per skill** — every skill is GitHub/ClawHub-ready from day one
- **Semantic versioning for skills** — clear rules for MAJOR vs MINOR vs PATCH

**Evolution** — not just 0→1, but 1→N:
- **Human-AI collaborative iteration** — ongoing improvement with versioned upgrades

It is not about what Claw *can* do — it is about turning an unbounded capability into a **reliable, high-quality pipeline**. The process is the product.

---

## Quick Start

Just send Claw an idea:

```
New skill: [name], [1-2 sentences], [source URL (optional)]
```

Claw enters **Phase 0** immediately — no setup required.

---

## Example: Building the `harness` Skill

Here is how Skill Factory handled a real request end-to-end:

**Input** (30 seconds):
```
New skill: harness, distill OpenAI's agent engineering article,
https://openai.com/index/harness-engineering/
```

**Phase 0** — Three questions answered in one message:
- Target: any Claw working on agent-first engineering
- Source: OpenAI Harness Engineering blog post
- Scope: GitHub + local workspace

**Phase 1** — `web_fetch` the source, saved to `wip/harness/raw/`

**Phase 2** — Distilled into `SKILL.md` (10 principles) + `references/` (two documents). Mandatory re-read caught 4 missed concepts including the "Ralph Wiggum loop".

**Phase 3** — Independent repo, MIT license, bilingual README, git-tagged `v1.2.0`, synced to workspace.

**Total time**: ~10 minutes. Without the factory, the same work took a full day of iteration and still had gaps.

---

## Four-Phase SOP Overview

| Phase | Name | Outpu

_meta.json

{
  "ownerId": "kn75mmeqmz6rt3hwpt8qy83cj1858fnw",
  "slug": "skillfactory",
  "version": "2.4.0",
  "publishedAt": 1778670169661
}

CHANGELOG.md

# Changelog — skill-factory

All notable changes to skill-factory will be documented in this file.

Format: [Keep a Changelog](https://keepachangelog.com/en/1.0.0/)

---

## [Unreleased]

No unreleased changes yet.

---

## [2.4.0] — 2026-05-13

### Summary

**End-to-end ClawHub CLI publishing workflow.** Phase 3 now includes a full, battle-tested SOP for publishing skill updates to ClawHub via CLI — covering login, publish, verify, and local sync. Lessons from earlier failures (proxy issues, requirement creep, device flow) are captured.

### Added
- Phase 3 → ClawHub CLI Publishing: device flow login, publish command, verify, local sync steps
- Phase 3 → Windows proxy notes: HTTPS vs SSH guidance
- Lessons Learned table: LRN-005 (ClawHub device flow background), LRN-006 (SSH/proxy), LRN-007 (requirement creep)

### Changed
- SKILL.md version bumped from v2.3.0 to v2.4.0
- README.md / README_zh.md version badges updated to v2.4.0

### Fixed
- CLAWHUB.md version updated from v2.1.0 to v2.3.0 (was out of sync)
- SKILL.md Workspace Layout: corrected `D:/WorkBuddy/` → `D:/Agent/WorkBuddy/skills/` for skill repo paths
- SKILL.md Legacy section: `D:/WorkBuddy/<skill-name>/` → `D:/Agent/WorkBuddy/skills/<skill-name>/`
- README_zh.md: OpenAI URL changed from Chinese-localized to canonical English URL
- README.md / README_zh.md: manual install path from `~/.claw/skills/` → `~/.workbuddy/skills/`
- .workspace-index.md: directory structure comment updated from v2.2.0 to v2.3.0

## [2.3.0] — 2026-05-12

### Summary

**First-class upgrade workflow for existing skills.** Skill Factory now explicitly supports 1→N maintenance alongside 0→1 creation.

### Added
- `Upgrade skill:` as a first-class entry point alongside `New skill:`
- Upgrade Workflow section in SKILL.md: Inspect → Classify → Propose → Apply → Summary
- Version bump decision framework: skill's "public API" defined (use cases, input/output, workflow phases, safety boundaries, file structure)
- Upgrade plan template: current version, proposed version, change type, files to modify, risk level, changelog entry, commit message, tag
- Enhanced trigger words covering upgrade and maintenance scenarios

### Changed
- SKILL.md: `[Unreleased]` section added to CHANGELOG guidance per Keep a Changelog standard
- CHANGELOG discipline: now requires grouped changes (Added / Changed / Fixed / Deprecated / Removed / Security) and user-readable summaries
- README.md: added "Maintaining Existing Skills (1→N)" section
- README_zh.md: updated quick start with skill upgrade example
- Legacy "Skill Iteration" section renamed to "Skill Iteration (Legacy)" with redirect to Upgrade Workflow

### Fixed
- Workspace layout path in SKILL.md now consistent with actual directory structure

---

## [2.2.2] — 2026-04-27

### Summary

**Version sync: local repo aligned with ClawHub v2.2.2 release.**

### Changed
- `SKILL.md`: version field updated to `2.2.2`
- `REA

CLAWHUB.md

# Skill Factory — ClawHub Listing

**Slug:** `skillfactory`  
**Display Name:** Skill Factory  
**Version:** 2.3.0  
**License:** MIT  
**Author:** PandaBro666

---

## One-Line Pitch

The machine that builds skills — because making a skill from scratch shouldn't feel like reinventing the wheel every time.

---

## What It Does

Skill Factory is your skill-making buddy. You bring an idea (or a URL), it handles the rest:

- **Lock the scope** — 3 quick questions to make sure you're building the right thing
- **Gather the goods** — fetches source material automatically
- **Distill the signal** — turns noise into structured SKILL.md + references
- **Ship it** — packages everything into a versioned, GitHub-ready repo
- **Keep it fresh** — structured upgrade workflow so your skills don't rot at v1.0

No boilerplate templates. No "what now?" moments. Just say the word and it works.

---

## When to Use

- You have a skill idea but don't know how to start
- You want to turn scattered notes or bookmarks into a proper skill
- You need to bring an existing skill up to date with new sources
- You're preparing something for GitHub or ClawHub
- You have a skill sitting at v1.0 and want to grow it without messing up

---

## Quick Start

Just tell your Claw:

```
New skill: [name], [1-2 sentences], [source URL (optional)]
```

Example:
```
New skill: prompt-engineering, best practices from Anthropic and OpenAI
```

Or to upgrade an existing one:
```
Upgrade skill: [name], [what to add], [new source URL (optional)]
```

---

## Tags

`meta-skill` `skill-authoring` `openclaw` `agent-engineering` `workflow-automation` `knowledge-distillation`

---

## Links

- **Repository:** https://github.com/zhelunSun/skill-factory
- **English README:** [README.md](README.md)
- **中文说明:** [README_zh.md](README_zh.md)
- **Changelog:** [CHANGELOG.md](CHANGELOG.md)

Github ReposUpdated 2d agoRank 70

AionUi

Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!

MCPOPENCLAW

Github ReposUpdated 6mo agoRank 70

activepieces

AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents

OPENCLAW

Github ReposUpdated 6mo agoRank 70

cherry-studio

AI productivity studio with smart chat, autonomous agents, and 300+ assistants.