# The Termi Protocol

> A 3D simulation of AI coding agent workflows. You run real CLI coding agents (Claude Code, Codex, Gemini CLI, Copilot, Aider and more) on your own machine; The Termi Protocol gives each one a body, a desk and a room, so every file read, command and change is visible live in 3D, with checkpoints, task boards, per-agent memory and live cost. macOS and Windows. One-time lifetime license, no subscription.

Every page below has a markdown twin: append `.md` to the path, or request the page with `Accept: text/markdown`.

## Product

- [Home](https://termiprotocol.com/index.md): what the product is, how it works, pricing summary
- [Download](https://termiprotocol.com/download.md): macOS and Windows builds
- [Pricing](https://termiprotocol.com/pricing.md): one-time lifetime license
- [Roadmap](https://termiprotocol.com/roadmap.md): nine protocols, one world
- [Chronicle](https://termiprotocol.com/news.md): release notes and news
- [Leaderboard](https://termiprotocol.com/leaderboard.md): global XP league

## Guides (Learn)

- [Learn hub](https://termiprotocol.com/learn.md): index of all guides
- [What is vibe coding?](https://termiprotocol.com/what-is-vibe-coding.md): definition, Karpathy origin, how it works, risks, staying in control
- [AI coding agents, explained](https://termiprotocol.com/ai-coding-agents.md): the agent loop, strengths, limits, supervision
- [The best AI coding agents in 2026](https://termiprotocol.com/best-ai-coding-agents.md): Claude Code, Codex CLI, Gemini CLI, Copilot, Aider, opencode, local models, compared from daily use
- [How to start vibe coding](https://termiprotocol.com/how-to-vibe-code.md): seven-step beginner setup, prompt templates, common mistakes
- [AI vibe coding tools in 2026](https://termiprotocol.com/vibe-coding-tools.md): the four-layer stack: CLI agents, AI editors, app builders, visibility
- [Running multiple AI coding agents in parallel](https://termiprotocol.com/multi-agent-ai-coding.md): worktrees, file locks, task boards, cost control
- [How to recover a closed Claude Code session](https://termiprotocol.com/recover-claude-code-session.md): claude --continue / --resume, plus Codex, Gemini CLI, Aider, opencode recovery

## Blog (founder notes by Omer Ercan)

- [Blog hub](https://termiprotocol.com/blog.md): all posts, newest first
- [Watching AI write code changed how I review it](https://termiprotocol.com/watching-ai-write-code.md): trajectory-first review, where agent risk hides in diffs
- [Six AI agents, one codebase](https://termiprotocol.com/six-agents-one-codebase.md): lanes, the glance economy, cost as a management tool, checkpoint review
- [Managing AI coding agents is becoming a real job](https://termiprotocol.com/managing-ai-coding-agents.md): the brief/delegate/inspect/unblock loop and which skills transfer
- [A week of barely supervised AI agents](https://termiprotocol.com/a-week-of-unsupervised-ai-agents.md): what broke (scope creep, weakened test, dep drift) and what surprised
- [Vibe coding is not no-code](https://termiprotocol.com/vibe-coding-is-not-no-code.md): constraint vs flood, ownership, the ladder of stakes
- [Why I built a 3D room for my AI coding agents](https://termiprotocol.com/why-i-built-a-3d-room-for-ai-agents.md): the founding story of The Termi Protocol
- [Omer Ercan](https://termiprotocol.com/omer-ercan.md): author page, founder bio and links
- RSS feed: https://termiprotocol.com/feed.xml

## Comparisons

- [Claude Code vs Cursor](https://termiprotocol.com/claude-code-vs-cursor.md): terminal agent vs AI editor, pricing, using both together
- [Cursor vs The Termi Protocol](https://termiprotocol.com/cursor-vs-termi-protocol.md): AI editor vs 3D agent workspace, where each wins, using both, pricing
- [Claude Code vs Codex CLI](https://termiprotocol.com/claude-code-vs-codex.md): repo awareness vs surgical speed, sandboxing, running both
- [Claude Code vs Gemini CLI](https://termiprotocol.com/claude-code-vs-gemini-cli.md): paid reliability vs the best free tier
- [Cursor vs VS Code with Copilot](https://termiprotocol.com/cursor-vs-vscode.md): completion, agent modes, pricing, fork trade-offs
- [Claude vs ChatGPT for coding](https://termiprotocol.com/claude-vs-chatgpt-for-coding.md): Fable and Opus against the GPT-5 series
- [Cursor alternatives](https://termiprotocol.com/cursor-alternatives.md): Windsurf, Zed, VS Code, Antigravity, terminal agents, free stacks
- [Claude Code alternatives](https://termiprotocol.com/claude-code-alternatives.md): Codex CLI, Gemini CLI, Aider, opencode, plus free, open-source and local-LLM picks
- [GitHub Copilot alternatives](https://termiprotocol.com/github-copilot-alternatives.md): Cursor, Windsurf, Cody, Tabnine, Supermaven, and terminal agents for autonomy
- [Lovable alternatives](https://termiprotocol.com/lovable-alternatives.md): Bolt.new, v0, Replit Agent, Bubble, and owning your stack with real CLI agents
- [Replit alternatives](https://termiprotocol.com/replit-alternatives.md): VS Code, Cursor, GitHub Codespaces, Gitpod, Bolt.new, and local agents you own
- [Windsurf alternatives](https://termiprotocol.com/windsurf-alternatives.md): Cursor, VS Code with Copilot, Zed, Antigravity, terminal agents, free stacks
- [ChatGPT alternatives for coding](https://termiprotocol.com/chatgpt-alternatives-for-coding.md): Claude, Codex CLI, Claude Code, Gemini, and agents that edit your repo directly

## Legal

- [Privacy Policy](https://termiprotocol.com/privacy.md)
- [Terms of Use](https://termiprotocol.com/terms.md)