universal-ai-config

First example skill for add-skill tests

Second example skill with a non-kebab name

Validate uac's assumptions about supported AI tools (Claude Code, Copilot, Cursor, etc.) against current upstream documentation and produce a written report of drift — renamed fields, new events, deprecated features, schema changes — so the user can plan follow-up changes. Does not modify source code.

Import existing AI tool configurations (from Claude, Copilot, Cursor, or Codex) into universal-ai-config templates. Converts target-specific files into universal templates.

Draft a pull request description from either the current chat session or the full feature branch (committed + uncommitted). Use when the user asks to write a PR description, prep a PR, or summarize changes for review.

Create, update, or manage universal-ai-config agent templates. Handles finding existing agents, deciding whether to create or modify, and writing the template.

Create, update, or manage AI configuration templates. Analyzes what the user needs and delegates to the appropriate template-specific skill.

Create, update, or manage universal-ai-config hook templates. Handles finding existing hooks, deciding whether to create or modify, and writing the template.

Create, update, or manage universal-ai-config instruction templates. Handles finding existing instructions, deciding whether to create or modify, and writing the template.

Create, update, or manage universal-ai-config MCP server templates. Handles finding existing configs, deciding whether to create or modify, and writing the template.

Create, update, or manage universal-ai-config skill templates. Handles finding existing skills, deciding whether to create or modify, and writing the template.

Scaffold a new target type (e.g., Zed, Windsurf) for universal-ai-config

Exercises all skill fields

Compress LLM instruction sets, system prompts, rules, and guidelines to reduce token count while preserving behavioral intent. Use this skill whenever the user wants to shorten, truncate, optimize, or compress prompts, system instructions, CLAUDE.md files, custom instructions, rules, guidelines, or any instruction set meant for an LLM. Also trigger when users mention "too many tokens," "prompt is too long," "reduce prompt size," or ask to make instructions more concise.

Generate tests for code

claude: Claude test generation skill

Create a new git release with changelog generation and GitHub release