parallel-dev

Coordinate backend and frontend AI subagents using an OpenAPI contract.

1|Updated Apr 1, 2026
One-click install
npx skills add https://github.com/zeptillionairplex/karpathy-claude-code-prompt-principle --skill parallel-dev-zeptillionairplex
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: parallel-dev
Source: https://github.com/zeptillionairplex/karpathy-claude-code-prompt-principle/tree/main/.claude/skills/parallel-dev
Command: npx skills add https://github.com/zeptillionairplex/karpathy-claude-code-prompt-principle --skill parallel-dev-zeptillionairplex

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires oapi-codegen, openapi-typescript, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill addresses the challenge of managing parallel project development of backend and frontend components driven by AI agents, ensuring consistency through explicit contracts.

Core Features & Use Cases

  • Contract-Centric Development: Uses an OpenAPI specification in contracts/openapi.yaml as the single source of truth for backend and frontend code generation.
  • Parallel Subagent Workflow: Enables simultaneous implementation by separate AI agents for backend and frontend based on layered rules and clear project structure.
  • Use Case: Initiate a project with an existing OpenAPI spec, spawn subagents in both backend and frontend directories, and coordinate their updates without drift.

Quick Start

Read the planning checklist in docs/04-planning-checklist.md, then generate the initial registry from docs/05-tech-stack-registry.md before starting subagent development.

Frequently Asked Questions about parallel-dev

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I coordinate parallel AI development for frontend and backend from a single OpenAPI spec?

Parallel AI development is coordinated by using an OpenAPI specification as the single source of truth, orchestrating subagents to generate backend and frontend code simultaneously without drift. This ensures spec compliance and consistent project restructuring.

What is the best way to manage API contract drift when using multiple AI agents for coding?

Managing API contract drift is achieved through contract-centric development, where an OpenAPI specification governs code generation for both frontend and backend. Layered rule enforcement and coordinated version control ensure all subagents remain synchronized.

Can I use oapi-codegen and openapi-typescript to orchestrate AI subagents across separate project layers?

Yes, you can use oapi-codegen and openapi-typescript to generate typed code from your OpenAPI spec, enabling separate AI subagents to implement backend and frontend layers in parallel within a structured, spec-first workflow.

How to start a spec-driven parallel workflow for multi-team AI-driven coding?

To start a spec-driven parallel workflow, read the planning checklist, generate the initial tech stack registry from your OpenAPI contract, and then spawn subagents in the respective backend and frontend directories to begin coordinated development.

Does contract-first AI project orchestration support scalable feature deployment across separate repositories?

Contract-first AI project orchestration supports scalable feature deployment by managing API contracts and enforcing layered rules across a layered repo structure, enabling seamless multi-team collaboration and iteration without versioning conflicts.

When should I not use a contract-first subagent workflow for AI-driven development?

A contract-first subagent workflow may not suit projects lacking a defined API structure, as it explicitly relies on an OpenAPI specification to drive simultaneous backend and frontend code generation and enforce spec compliance.