Traction scoreGitHub stars can be faked, so popularity alone can be misleading. Traction Score looks for broader signs of recent attention, adoption, and active maintenance.
This collection provides structured rulebooks that align AI agent outputs with established software engineering principles. It helps technical leads and architects ensure that generated code adheres to rigorous standards for module boundaries, complexity reduction, and system scalability.
By integrating specific frameworks like Clean Architecture and Domain-Driven Design, the repository transforms AI generation from unpredictable drafting into a disciplined process that separates business policy from infrastructure details.
WHO IT'S FOR
Technical leads / architects
standardizing AI agent coding patterns across teams
Backend / infrastructure engineers
designing scalable and reliable data systems with AI
Software engineers using AI assistants
improving code readability and maintainability via agents
Platform / DevEx teams
creating shared AI prompt libraries for software design
Compatible AgentsThe repository documents support for these agents. The skills may also work with other agents that can load SKILL.md files, but they may need some setup or small changes.
Optional description of the behavior that must remain unchanged
Use the refactoring skill to review the existing code in <path-to-file-or-module>. Identify one concrete code smell and propose a small, behavior-preserving refactoring sequence. State which existing tests protect the behavior and which focused test should be added first if coverage is missing. Keep the refactoring separate from feature changes; do not modify files yet.
Brief description of the intended change or risky behavior
Use the working-effectively-with-legacy-code skill to inspect <path-to-legacy-code>. Suggest one characterization test that captures current behavior and identify one seam or dependency-breaking step that would make the intended change safer. Limit the result to a small first step and do not modify files yet.
Optional description of the dependency and expected traffic or failure conditions
Use the release-it skill to review the integration implemented in <path-to-integration-code>. Produce a short risk assessment covering its timeout, retry, circuit-breaker, backpressure, and observability behavior, then recommend the single highest-value improvement. Do not modify or deploy anything.
Claude
Codex
Cursor
Copilot