AI Agent Hub
Back to skills
🤖

Skill Creator v2.0

AI Agent Updated 2026.08.30

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

Please install @user_2dc8a2f8/guipi-skill-factory into my AI assistant by following https://skillhub.cn/install/skillhub.md.

About this skill

Problem

When maintaining multiple AI agent skills, new skills can overlap with existing ones in function, triggers, and data sources, causing routing conflicts and extra maintenance work. guipi-skill-factory changes skill creation from directly writing a SKILL.md into a workflow that first evaluates the existing skill library before deciding whether to integrate, upgrade, or create a new skill.

Workflow

  • Extract the new skill's core function, trigger scenarios, data sources/tools, and target domain, asking follow-up questions when the request is unclear.
  • Scan existing skills across functional, trigger, and data-source dimensions, with special attention to high-overlap areas such as content creation, data analysis, and compliance review.
  • If overlap is high, suggest integrating the feature into the original skill and incrementing the version. If overlap is partial but adds meaningful scope, suggest an enhanced version. If coverage is low, create a new skill and add a skill-system analysis section.
  • Run a security check before creation, flagging hard-coded API Key values, command injection, privacy leakage, over-permissive triggers, and dependency issues. P0 and P1 risks require user confirmation.
  • After creation or upgrade, optionally publish to Xiaping, a private GitHub repository, or ClawHub.

Boundaries

It fits personal or team workflows that already maintain a skill library and want consistent naming and versioning. Publishing depends on the corresponding platform skills; without them, creation and upgrade can still proceed. Avoid using new version or enhanced as the skill name; version changes should appear in the version number or changelog.

Use Cases

  • Before adding a new topic-evaluation agent skill, run a three-dimension overlap check against existing skills.
  • Upgrade an existing compliance-review skill by adding platform rules and keeping the version record in a new SKILL.md.
  • Audit a draft SKILL.md for API keys, shell injection, over-permissive triggers, and dependency risks before creation.
  • Finalize skill naming, versioning, and publish-ready output for Xiaping, private GitHub, or ClawHub.

Best For

  • Engineers maintaining a personal AI agent skill library and avoiding functional or trigger conflicts between skills.
  • Agent development leads who need to standardize scattered skill docs into SKILL.md with version records.
  • Independent developers who want to self-check keys, shell injection, and permission risks before publishing.
  • Agent ecosystem operators syncing skills to Xiaping, private GitHub, or ClawHub.