Praxis
New给「AI 驱动开发」立规矩的 Claude Code skill 方法论库:七层文档治理 · 对抗评审 · 任务总控三驾马车,外加老代码考古、施工蓝图等共 16 个 skill —— 让 AI 写代码又快又不失控。
Summary
Praxis is a comprehensive Claude Code skill that enforces structured AI-driven development through 16 specialized sub-skills, including seven-layer document governance, adversarial review, and task orchestration.
- It helps developers maintain control over AI-generated code by providing methodologies for legacy code archaeology, construction blueprints, and quality assurance, ensuring speed without chaos.
Install & Usage
mkdir -p .claude/skillsmkdir -p .claude/skills && curl -o .claude/skills/praxis.md https://raw.githubusercontent.com/BackToCimaCoppi/Praxis/main/SKILL.md/praxisUse Cases
Usage Examples
/praxis blueprint design a microservice for user authentication with JWT and OAuth2 support
/praxis legacy-archaeology analyze the src/ directory and produce a dependency graph and refactoring roadmap
/praxis adversarial-review review the changes in the current branch for security vulnerabilities and logic errors
Security Audits
Frequently Asked Questions
What is Praxis?
Praxis is a comprehensive Claude Code skill that enforces structured AI-driven development through 16 specialized sub-skills, including seven-layer document governance, adversarial review, and task orchestration. It helps developers maintain control over AI-generated code by providing methodologies for legacy code archaeology, construction blueprints, and quality assurance, ensuring speed without chaos.
How to install Praxis?
To install Praxis: create the skills directory (mkdir -p .claude/skills), then run: mkdir -p .claude/skills && curl -o .claude/skills/praxis.md https://raw.githubusercontent.com/BackToCimaCoppi/Praxis/main/SKILL.md. Finally, /praxis in Claude Code.
What is Praxis best for?
Praxis is a skill categorized under General. Created by BackToCimaCoppi.
What can I use Praxis for?
Praxis is useful for: Establish a seven-layer document hierarchy to guide AI code generation with clear context and constraints.; Run adversarial code reviews where AI critiques its own output to catch bugs and design flaws before merge.; Orchestrate complex multi-step development tasks by breaking them into subtasks with dependencies and validation gates.; Analyze legacy codebases to generate structured documentation and refactoring plans for AI-assisted modernization.; Create detailed construction blueprints that specify architecture, APIs, and data flow before any code is written.; Enforce coding standards and best practices automatically by applying rule-based checks during AI code generation..