ai-pair

Coordinate developer and two reviewer AI models for parallel task reviews.

335|59|Updated Mar 8, 2026
One-click install
npx skills add https://github.com/axtonliu/ai-pair --skill ai-pair
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ai-pair
Source: https://github.com/axtonliu/ai-pair/tree/main
Command: npx skills add https://github.com/axtonliu/ai-pair --skill ai-pair

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Coordinating heterogeneous AI models to collaboratively work on a single task, enabling diverse perspectives to review and refine outputs.

Core Features & Use Cases

  • Role-based team orchestration: assign a developer/author and two reviewers (Codex and Gemini) to a task.
  • Semi-automatic workflow: user stays in control while agents execute actions and report back.
  • Multi-domain applicability: supports code, articles, newsletters, scripts, and other creative outputs.
  • Cross-model reviews: parallel analysis by Codex and Gemini to improve quality and maintainability.

Quick Start

Start a dev-team with /ai-pair dev-team [project] to begin the multi-model review workflow.

Frequently Asked Questions about ai-pair

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

FAQPage Schema
How do I coordinate multiple AI models to review my code?

Multi-model code review is coordinated by assigning a developer role to one model and reviewer roles to two others, enabling parallel analysis to refine and improve outputs collaboratively.

What do I need to set up before running a multi-model AI review workflow?

Running a multi-model AI review workflow requires installing and authenticating Claude Code, Codex CLI, and Gemini CLI to enable the heterogeneous agents to execute actions and report back.

Can I use multi-model agent teams to review articles and newsletters?

Multi-model agent teams support articles, newsletters, scripts, and other creative outputs by assigning an author role and two reviewer roles to collaboratively refine content quality.

How do I start a multi-model review session from the CLI?

Start a multi-model review session by running the /ai-pair dev-team command with your project name to launch the role-based orchestration workflow for parallel code or content analysis.

Do I stay in control when AI agents execute actions during the review?

The semi-automatic workflow keeps you in control by letting you decide when to launch reviews, while the AI agents execute assigned actions and report findings back for your evaluation.

What are the limitations of using heterogeneous AI models for collaborative tasks?

Heterogeneous AI model collaboration requires authenticated access to multiple CLI tools simultaneously and limits execution to semi-automatic workflows where user intervention is needed to trigger reviews.