Yardlet

Yardlet is a Mac application designed for developers who use multiple coding agent command-line interfaces (CLIs) such as Codex and Claude Code. It enables users to run various agent CLIs side by side, each maintaining its own authenticated session and task history. This means developers can manage and switch between their preferred code assistants without losing session context or workflow, streamlining their workflow on a single Mac dashboard.

The app features a unified interface with consistent controls, conversation views, and approval management for each agent, regardless of the underlying model. Developers can initiate, review, and merge tasks across different agents, and even have one agent review the output of another. Key tools provided by Yardlet include messaging between agents, launching subtasks across different agents, and tracking session limits per account. The system enforces clear approvals by recording and transmitting user instructions verbatim, maintaining transparency and control when delegating work across agents.

Being open source and distributed under the MIT License, Yardlet is free to install and safe to try without mandatory third-party connections. It's purpose-built for macOS (Apple silicon, macOS 15+), highlighting a focus on performance and privacy for developers comfortable with local tools. The app’s support for instant task handoff and collaboration between code assistants makes it particularly valuable for users working on complex projects that benefit from diverse agent capabilities.

Yardlet is intended for software developers, engineers, and technical leads who actively use AI coding agents like Codex and Claude Code within their daily workflow. It's ideal for those managing multiple AI assistants on macOS and looking for a way to keep different agent sessions separated, organized, and readily accessible in one place.

Technical teams in startups or small companies that experiment with or depend on multiple code-focused AI models will benefit from Yardlet's capability to run agents in parallel, approve merge requests, and delegate coding tasks or reviews between agents. It's especially useful for professionals who value open-source solutions, prefer local Mac apps over web-based interfaces, and need a transparent approval workflow to track task delegation and agent collaboration on software projects.