EveryInc / compound-engineering-plugin

An AI coding plugin that compounds engineering knowledge across development cycles, making each unit of work easier than the last through structured planning, review, and knowledge codification.

活跃维护 MIT TypeScript Tracked
24.7k 2.0k 5 小时前
CINPM
compound engineering

星标趋势

数据积累中,暂无足够数据生成趋势图

AI 分析

项目摘要

A TypeScript plugin providing AI-powered engineering skills for Claude Code, Codex, Cursor, and other AI coding assistants. It implements a 'compound engineering' philosophy where each development cycle—brainstorm, plan, work, simplify, review, compound—makes subsequent work easier through structured workflows and knowledge codification.

为什么值得关注

With nearly 23,000 stars and 100 releases in 6 months, this project has rapidly gained traction for its structured approach to AI-assisted development. Its core innovation is the compounding loop: learnings from each cycle are codified into reusable knowledge that grounds future brainstorming and planning, creating a flywheel effect that improves code quality and team velocity over time.

优势

  • Well-articulated philosophy with clear 80/20 planning-to-execution ratio
  • Active development with 100 releases in 6 months and 69 contributors
  • Multi-platform support across Claude Code, Codex, and Cursor
  • Knowledge compounding loop that improves future AI interactions
  • Strong CI/CD practices with tests and automated workflows

局限性

  • No examples directory to help new users get started quickly
  • No Docker support for containerized development environments
  • Relatively new project (created Oct 2025) with evolving API surface
  • Tied to specific AI coding platforms, limiting standalone utility

使用场景

  • Structured AI-assisted feature development with built-in planning and review
  • Team knowledge management through codified learnings in docs/solutions/
  • Multi-agent code review workflows that catch patterns, not just bugs
  • Onboarding acceleration by compounding institutional knowledge for AI agents
目标用户: Software engineering teams and individual developers using AI coding assistants who want disciplined, repeatable workflows that improve over time rather than one-off prompt interactions
学习曲线:
分析模型:LongCat-2.0 | 分析时间:1 个月前