Agent Clan

Product Introduction

1. Definition: Agent Clan is an independent desktop workspace application designed for Mac and Windows operating systems that visually renders AI coding agents—specifically Claude Code and Codex—as interactive 3D characters on a user's screen. Functioning as a graphical user interface (GUI) overlay layer, the software combines local computer vision (hand tracking) and speech-to-text processing to orchestrate software engineering tasks.
2. Core Value Proposition: Agent Clan exists to streamline AI-assisted software development by replacing traditional terminal switching and manual prompt copying with spatial computing gestures, voice commands, and persistent desktop visualization. By offering an "agent army" interface, it maximizes user workflow efficiency ("AI aura") for managing multiple asynchronous autonomous coding workflows simultaneously.

Main Features

1. Spatial Gesture and Voice Control Engine: Agent Clan processes hand movements and voice inputs locally on the user's machine using built-in computer vision and microphone integration. Users can trigger agent interfaces by making a fist, select specific 3D avatars via pinch-and-pull gestures, and convert spoken natural language into actionable coding tasks without using a physical keyboard or mouse.
2. Always-On-Top 3D Workspace & Heads-Up Display (HUD): The application renders avatars in a 3D virtual room while maintaining persistent floating indicator heads over all third-party desktop applications. Visual status rings indicate whether an agent is actively computing, has completed a task, or requires user intervention. Permission gates (allow/deny) are managed via intuitive directional pulls (right to allow, left to deny).
3. Drag-and-Drop Context Pipelines and Pro Customization Mods: Users can drop files, images, or raw text snippets directly onto floating agent heads to establish instant prompt contexts. Under the Pro tier, users can leverage Anthropic Claude agents to write custom software modifications ("mods") using Agent Clan's Model Context Protocol (MCP) tools, enabling real-time UI/UX customizations and feature extensions without restarting the application.

Problems Solved

1. Pain Point: Developers managing multiple Claude Code or Codex sessions often experience context switching fatigue, terminal window clutter, and friction in handling background permissions and prompt distribution.
2. Target Audience: Software engineers, full-stack developers, DevOps professionals, and technical project managers who utilize autonomous AI coding tools and require streamlined multi-agent orchestration.
3. Use Cases: Executing parallel terminal processes (e.g., running npm install, managing repository refactors, or debugging errors) across multiple codebase branches while monitoring agent execution status from a single consolidated desktop interface.

Unique Advantages

1. Strengths & Limitations (Pros & Cons):
   • Pros: Completely local hand-tracking and voice transcription privacy (no camera frames uploaded to cloud servers); seamless multi-agent orchestration over any native desktop app; intuitive gesture-based permission handling; support for infinite parallel terminal desks under the Pro tier.
   • Cons: Gesture and voice processing require functional local hardware specifications (webcam and microphone); AI service usage fees are billed separately by underlying providers (Anthropic/OpenAI); Codex agents lack native mod-writing capabilities compared to Claude integration; camera tracking performance relies on ambient desk lighting.
2. Key Alternatives & Differentiation:
   • Traditional Terminal Tabs (iTerm2, Windows Terminal): Unlike text-based terminals, Agent Clan offers visual 3D spatial awareness, gesture control, and persistent HUD status rings for asynchronous agent management.
   • Web-based AI Dashboards: Unlike browser interfaces that require manual context copying, Agent Clan features system-wide drag-and-drop prompt inputs and native desktop application integration.
   • IDE Extensions (VS Code Copilot Chat, Cursor): While IDE plugins operate strictly within a single code editor environment, Agent Clan functions as a cross-application desktop overlay layer capable of orchestrating external command-line tools like Claude Code and Codex.

Frequently Asked Questions (FAQ)

1. Question? Does Agent Clan replace my existing Claude Code or Codex accounts? Direct, keyword-rich answer: No. Agent Clan is an independent desktop workspace that connects to your existing Claude Code and Codex accounts. You continue to use your own API accounts, and the original AI service providers bill usage separately.
2. Question? Are my camera video feeds and audio recordings uploaded to the cloud? Direct, keyword-rich answer: No. Gesture recognition and speech-to-text transcription run entirely locally on your computer. Agent Clan does not upload raw camera frames or audio recordings; only the final transcribed text is sent to your chosen AI agent.
3. Question? How does the Pro tier lifetime update model work? Direct, keyword-rich answer: The Pro tier is available as a one-time purchase (priced at $14 during launch) which grants unlimited projects, agents, and custom mod capabilities, along with 1-year of software updates applicable across up to 3 computers linked to your account.
4. Question? Can I use Agent Clan without hand gestures or a webcam? Direct, keyword-rich answer: Yes. While Agent Clan supports spatial hand tracking and voice input, all features—including agent selection, prompt submission, and permission handling—are fully accessible using a standard computer mouse and keyboard.