Termi Protocol

Product Introduction

1. Definition: The Termi Protocol is a local-first, 3D visualization and management platform for AI coding agents. It is a desktop application that functions as a visual command center, simulating AI agent workflows in a real-time 3D environment.
2. Core Value Proposition: It exists to solve the opacity and lack of control in AI-assisted coding by transforming abstract terminal logs into an immersive, observable 3D workspace. Its primary value is providing full visibility, persistent context, and granular control over autonomous coding agents, turning a blind trust process into a transparent, manageable simulation.

Main Features

1. Live 3D Agent Simulation: This is the core visualization engine. It renders AI agent actions (reading files, writing code, executing shell commands) as real-time animations within a 3D room. How it works: The application interfaces with your local CLI agent (e.g., Claude Code, Aider), parses its standard output and file system interactions, and maps them to predefined animations and environmental interactions for a specific agent "body" at its desk.
2. Integrated Command Center: A multi-panel control dashboard per agent featuring a genuine, fully interactive terminal shell (built on xterm.js and a PTY), a live cost tracker (token/dollar consumption), a Kanban-style task board the agent manipulates, a checkpoint rewind system, and persistent project memory storage. This consolidates all agent telemetry and control surfaces into a single view.
3. Persistent Project State & Multi-Agent Coordination: The platform automatically saves task boards, code checkpoints, and each agent's learned context ("project memory") between sessions. It implements file-locking mechanisms to prevent collisions when multiple agents are working concurrently, enabling safe multi-agent workflows with handoff capabilities.
4. Gamified Productivity Layer (Moscow Protocol): A meta-layer that awards experience points (XP) for real coding workflow milestones. Users level up from Rookie to Legend, unlocking ranks and competing on a global leaderboard. This layer is designed to incentivize and benchmark efficient AI agent utilization.
5. Customizable Workspace & Notification System: Includes a room editor for arranging desks and focus tools, and a pet system that evolves with user activity. The alert system provides audible and desktop notifications when an agent requires human approval, enabling asynchronous, interrupt-driven supervision.

Problems Solved

1. Pain Point: The "black box" nature of AI coding agents. Developers cannot intuitively see or understand the step-by-step process an agent takes, leading to blind approvals, lost context after closing a terminal, and difficulty debugging when agents "go off the rails."
2. Target Audience: Software Engineers and Developers who regularly use CLI-based AI coding assistants (Claude Code, Aider, Cursor, etc.); Tech Leads managing or prototyping with multiple AI agents; Indie Hackers and Solo Founders leveraging AI for full-stack development who need oversight without constant monitoring.
3. Use Cases: Code Review & Agent Debugging: Visually stepping through an agent's proposed changes before approval. Long-Running Project Development: Maintaining persistent context and task state across multiple work sessions. Multi-Agent Orchestration: Safely coordinating specialized agents (e.g., one for frontend, one for backend) within a single project with conflict prevention. Educational Demonstration: Teaching teams or stakeholders how AI development agents operate in a tangible, engaging format.

Unique Advantages

1. Differentiation: Unlike standard terminal emulators or IDEs with AI plugins, Termi Protocol does not just display logs—it simulates the workflow in a spatial, game-like environment. Unlike screen recording or log replay tools, it offers real-time interaction, control, and state persistence. Its local-first architecture and use of your own API keys differentiate it from cloud-based AI platforms that host your code.
2. Key Innovation: The 3D spatial metaphor for DevOps observability. By mapping file I/O to an agent walking to a cabinet, shell commands to terminal interaction at the desk, and code writing to typing animations, it creates a novel, intuitive mental model for understanding complex, automated processes. The integration of a real PTY-backed terminal within the 3D simulation ensures it is not a passive visualization but a fully functional control interface.

Frequently Asked Questions (FAQ)

1. Is the Termi Protocol a subscription service? No, it is a one-time purchase lifetime license. The "Founding Backer" offer provides unlimited, permanent access to the current application (Milano protocol) and all future protocol chapters and features with no recurring subscription fees or usage caps.
2. How does Termi Protocol handle my code and API keys? The application is architected as a local-first client. Your AI agent (e.g., Claude Code) runs on your local machine using your own API keys. Termi Protocol acts as a visualizer and controller for that local process; your source code and credentials never leave your system unless explicitly configured to do so by your chosen CLI agent.
3. What do I need to run the Termi Protocol software? You need a compatible desktop operating system (currently macOS or Windows) and an existing AI coding CLI agent (such as Claude Code, Aider, or Gemini CLI) installed and configured with its own API keys. Termi Protocol visualizes the workflow of these external agents.
4. Can I use Termi Protocol with any AI model or agent? It is compatible with any command-line interface (CLI) based AI coding agent. The platform connects to the agent's standard input/output streams. Officially mentioned integrations include Claude Code, Codex, Gemini CLI, Aider, and Grok CLI, but the architecture supports "Any CLI agent."
5. What happens if I close the Termi app during an agent's task? The system is designed for persistence. Your agent's task board, the history of checkpoints (snapshots of changes), and the agent's project-specific memory are automatically saved. Upon reopening the application and the specific "room," you can resume the workflow from the last checkpoint or review the full history.