Three focused skills that help Claude Code and Codex reason about NestJS architecture, object design, framework features, performance, and scale—without forcing accidental complexity.
Each skill owns a distinct decision level. Use one as the lead and bring in another only when the task crosses a real boundary.
The skills share the same defaults: inspect the actual repository, preserve explicit contracts, prefer the least complex safe design, keep policy separate from volatile infrastructure, and verify outcomes with tests or measurements. Their scopes overlap at intentional handoff points—not through contradictory ownership.
A practical sequence
For a large change, decide the system boundary first, design the collaborating objects second, then choose the NestJS runtime features and performance controls that implement it.
The open Agent Skills format works with Claude Code and Codex.
npx skills add amirtaherkhani/nestjs-agent-skills