agent-skills vs compound-engineering-plugin

Structured engineering workflows for AI coding agents

agent-skills and compound-engineering-plugin both guide AI coding agents through structured engineering work. agent-skills emphasizes lifecycle coverage, verification, and shared checklists, while compound-engineering-plugin emphasizes capturing reusable lessons for future tasks.

agent-skillscompound-engineering-plugin
LanguageJavaScriptTypeScript
LicenseMITMIT
Stars109k25.4k
Forks11.3k2.1k
Last analyzedOct 20, 2026Oct 3, 2026

Key differences

  • agent-skills includes 25 skills, nine slash commands, specialist agent personas, and shared checklists; compound-engineering-plugin offers skills for planning, execution, review, and related coding tasks.
  • agent-skills highlights checkpoints and per-task verification, including in its `/build auto` workflow; compound-engineering-plugin highlights recording project-specific lessons and an autonomous workflow that can proceed through review and shipping steps.
  • agent-skills supports native integrations or portable Markdown skill files; compound-engineering-plugin supports hosts including Claude Code, Cursor, and Codex, with installation and invocation varying by host.
  • agent-skills is written in JavaScript, while compound-engineering-plugin is written in TypeScript; both use the MIT license.
  • The provided project facts list 101k stars and 10.6k forks for agent-skills, compared with 25.4k stars and 2.1k forks for compound-engineering-plugin.

Choose agent-skills if you…

  • need lifecycle workflows covering requirements, planning, implementation, testing, review, and shipping.
  • want explicit verification checkpoints, specialist personas, and shared engineering checklists.
  • prefer portable Markdown skill files or want to introduce workflows incrementally.
Read the agent-skills analysis →

Choose compound-engineering-plugin if you…

  • want agents to record reusable engineering lessons for later tasks.
  • need skills for brainstorming, planning, code explanation, debugging, and Git workflows alongside implementation and review.
  • want an autonomous workflow that can take work from a plan or bug fix through review and shipping steps.
Read the compound-engineering-plugin analysis →

Compare the numbers side by side, or add more projects

This comparison is generated with AI from the OSRepos analyses of both projects. Always check each project's repository and documentation before choosing.