Xaction Skills — AI skill for Claude Code
8 production-grade agent skills (SKILL.md standard) by Xaction: code-review, debugging, tdd, file-organizer, markitdown, skill-builder, skill-reviewer, text-patterns.
How to install Xaction Skills
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open CarolZ1/xaction-skills and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Xaction Skills does
8 production-grade agent skills (SKILL.md standard) by Xaction: code-review, debugging, tdd, file-organizer, markitdown, skill-builder, skill-reviewer, text-patterns. / 8 个生产级 AI 智能体技能:代码审查、调试、TDD、文件整理、文档转换、技能制作/审查、文本处理模式。
Alternatives in AI
- Humanizer: Remove AI Writing Patterns — You are a writing editor that identifies and removes signs of AI-generated text to make writing sound more nat 10.5k ★
- Huashu Md HTML — md/html 双向流水线 · 三个能力一站式:万物→md / md→精美html / html→md 889 ★
- Claude Code Configs — A comprehensive collection of production-grade Claude Code configurations, specialized agents, and automation 624 ★
README
Xaction Skills
8 production-grade agent skills — cross-platform, ecosystem-agnostic, following the open SKILL.md standard.
8 个生产级智能体技能——跨平台、不绑定生态,遵循开放的 SKILL.md 标准。
**English** | [中文说明](#中文说明)
Skills
| Skill | What it does |
|---|---|
| code-review | Multi-dimensional code review with quality gates — five axes (correctness, readability, architecture, security, performance), severity-labeled findings (Critical / Required / Optional / Nit), and a clear approval standard |
| debugging | Four-phase systematic debugging with the Iron Law: no fixes without root-cause investigation first. Escalates to architecture questioning after 3 failed fixes |
| tdd | Test-driven development discipline — red → green → refactor loop, seams, anti-patterns, mocking principles, and worked examples for TS/Python/no-framework |
| file-organizer | Safely organize files and folders — read-only scan first, SHA-256 duplicate detection, strict safety guardrails, trash over delete |
| markitdown | Convert documents (PDF, Word, Excel, PPT, HTML, images, audio) to Markdown via Microsoft MarkItDown — safe local conversion, batch workflows, offline fallback |
| skill-builder | Build, modify, optimize, or review skills — a meta-skill with trigger-word engineering, structure guide, and quality checklists |
| skill-reviewer | Review a skill and score it automatically against an 11-dimension standard, with hard-check scripts and a P1/P2/P3 fix list |
| text-patterns | Curated text-processing patterns based on Daniel Miessler's fabric — summarize, extract insights, analyze claims, improve writing; works with or without the fabric CLI |
Installation
Each folder is a self-contained skill package. Drop the folder you want into your agent's skills directory:
# Claude Code
git clone https://github.com/CarolZ1/xaction-skills.git
cp -r xaction-skills/tdd ~/.c
Related Skills
Better Skill Builder
Skill that teaches AI agents how to build Agent Skills for Claude Code, Codex, Amp, and OpenCode. Production p
Unofficial Claude Code Prompt Playbook
Unofficial playbook for production-grade LLM system prompt architecture, derived from local analysis of Claude
Agent Blueprint
Build production-grade AI agents from scratch. Patterns, templates, and references for workflow builders and a
Ring
89 skills and 38 specialized agents that enforce proven engineering practices for AI-assisted development. TDD
Agents Md Snippets
Copy-paste AGENTS.md / CLAUDE.md snippets from my AI coding workflow videos: context re-entry, worktrees, TDD,
Ultimate Debugger Prompt For Claude 3
A professional-grade debugging AI prompt designed for Claude 3.5 Sonnet that meticulously examines errors, gen
Related Agents
Django Reviewer
Expert Django code reviewer specializing in ORM correctness, DRF patterns, migration safety, security misconfi
Python Document Processor
Production-grade Python document processor for PDF/DOCX/Markdown/TXT extraction with robust error handling, te
Conformance Reviewer
Review an IRIS Interoperability production against the iris-interop best-practice criteria AFTER it is built a