AI Agent Hub
Back to skills
Skill Factory Pro icon

Skill Factory Pro

AI Agent Updated 2026.08.30

Paste the following prompt into your AI chat to install this skill:

Please install @user_ad872d55/skill-factory-pro according to https://skillhub.cn/install/skillhub.md.

About this skill

Problem

Agent skills often drift across D:/Skills, .qclaw/skills, and workspace-main/skills, while whitelist references break after deletions or renames. New skills can duplicate existing ones, inaccurate description fields hurt routing, and long main files waste tokens. This skill turns skill creation into an auditable workflow.

How It Works

It proceeds by phases: three-question scoping defines trigger context, value, and monetization; existing-skill scanning decides whether to merge or create a new skill. The generation stage uses domain agents to fill knowledge gaps and outputs SKILL.md, references/, and packaging notes. Before delivery, it checks whether the description covers feature keywords and triggers, whether the main file stays compact, and whether a quality score meets the publish threshold.

Boundaries

It treats D:/Skills as the manual source, then reverse-syncs the other two directories and cleans whitelist entries. The governance module is read-only by default and requires confirmation before changes. If existing skills differ substantially, creating a new skill may be cleaner; for one-off use without publishing or shared maintenance, some monetization and sync steps may be heavier than needed.

Use Cases

  • Create a TCM consultation skill package through three-question scoping, agent research, and scoring.
  • Scan D:/Skills when a new idea overlaps an old skill, then merge modules instead of duplicating.
  • Run audit across agent skills to find stale whitelist entries and version drift in three directories.
  • Rewrite description and trigger before release to keep feature keywords and routing words under 120 characters.

Best For

  • Agent skill maintainers who need to prevent directory drift and stale whitelist references.
  • SkillHub package authors who need publishable SKILL.md files plus references.
  • Routing-quality owners who need consistent description and trigger keywords.
  • Local skill-library operators who want read-only audits of duplicates and stale references.