Zero: The AI Coding Agent for Your Local Terminal
This repository profile is provided by osrepos.com, an open source repository discovery platform.

Summary
Zero is an innovative AI coding agent designed for your local terminal, offering powerful capabilities to inspect repositories, edit files, run commands, and utilize browser/terminal helpers. It provides durable local sessions while giving users full control over the AI model and permission levels. This tool empowers developers with a customizable and secure AI assistant directly within their development environment.
Repository Information
Topics
Click on any tag to explore related repositories
Use at your own risk
OSRepos shares public repositories for knowledge and discovery only. Any installation, execution, configuration, or use of code from these repositories is the user's own responsibility. Always review the repository, source code, dependencies, licenses, and security implications before running or installing anything. OSRepos is not responsible for issues, damages, or losses resulting from third-party repositories.
Introduction
Zero is a sophisticated AI coding agent built for your local terminal, putting you in command of your development workflow. It allows you to inspect codebases, modify files, execute commands, and leverage browser/terminal utilities, all while maintaining persistent local sessions. With Zero, you choose your preferred AI model and define the permission levels, ensuring a secure and tailored experience. It's designed to be a powerful, extensible, and privacy-focused assistant for developers.
Installation
Getting started with Zero is straightforward. Here are the most common installation methods:
npm
The npm package provides a convenient wrapper that installs the appropriate Zero binary for your platform (Linux, macOS, Windows on x64 and arm64).
npm install -g @gitlawb/zero
zero
Install Scripts
For Linux/macOS, you can use a simple curl command:
curl -fsSL https://raw.githubusercontent.com/Gitlawb/zero/main/scripts/install.sh | bash
Windows users can install via PowerShell:
irm https://raw.githubusercontent.com/Gitlawb/zero/main/scripts/install.ps1 | iex
From Source
If you prefer to build from source, Go 1.25+ is required:
git clone https://github.com/Gitlawb/zero.git
cd zero
go run ./cmd/zero
For more detailed installation instructions, refer to the official documentation.
Examples
Zero offers both an interactive TUI (Terminal User Interface) and a headless exec mode for scripting.
Interactive TUI
Simply run zero to start the TUI. A setup wizard will guide you through selecting a provider and model.
zero
Common slash commands within the TUI include:
/model,/provider: Switch the active model or provider./spec,/plan: Draft and review a plan before building./image: Attach an image for vision-capable models./resume,/rewind: Continue or roll back local sessions.
Headless exec
For one-shot commands or scripting, use zero exec:
zero exec "explain internal/agent/loop.go"
zero exec --model claude-sonnet-4.5 "refactor the config loader"
zero exec --use-spec "add rate limiting to the API client"
Zero also supports stream-JSON for programmatic input/output, making it suitable for CI/CD pipelines.
Why Use Zero
Zero stands out for several compelling reasons:
- Model Flexibility: Use your preferred AI model, including OpenAI, Anthropic, Gemini, Groq, OpenRouter, DeepSeek, Mistral, xAI, Qwen, Kimi, GitHub Models, Ollama, LM Studio, or any OpenAI/Anthropic-compatible endpoint.
- Full Control: Maintain complete control over file writes, shell commands, network access, and out-of-workspace writes through Zero's robust permission and sandbox policy.
- Terminal-Native Experience: The TUI provides a rich interactive experience with model/provider pickers, image input, slash commands, live plan/tool rendering, scrollback, themes, and session management.
- Scriptability: The
zero execcommand is highly scriptable, supporting text/JSON/stream-JSON I/O, isolated worktrees, spec-first runs, and meaningful exit codes for CI environments. - Local Context: All sessions are stored locally on disk, are searchable, resumable, and never uploaded as telemetry, ensuring your privacy.
- Extensibility: Extend Zero's capabilities with MCP servers, skills, plugins, hooks, and specialist subagents, all accessible from the same CLI.
Links
- GitHub Repository: https://github.com/Gitlawb/zero
- Installation Documentation: docs/INSTALL.md
- Stream-JSON Protocol: docs/STREAM_JSON_PROTOCOL.md
Related repositories
Similar repositories that may be relevant next.

AgentSkills: A Curated Collection for LLM Agent Skills and Resources
August 20, 2026
AgentSkills is an extensive curated collection of resources, papers, tools, projects, and frameworks focused on building and deploying skills for large language models. This repository serves as a central hub for understanding the LLM skills ecosystem, from Anthropic's official systems to academic research and open-source agent frameworks. It is an invaluable resource for anyone exploring the rapidly evolving field of AI agents.

mcp-gateway: Unifying AI Tool Access with Reduced Context Overhead
August 15, 2026
mcp-gateway is a powerful Rust binary designed to streamline AI agent interaction with diverse tools. It consolidates unlimited MCP servers and REST APIs behind a single, compact endpoint, drastically reducing context token overhead and enabling efficient tool access.

Jan: An Open-Source, Offline ChatGPT Alternative for Your Desktop
August 14, 2026
Jan is a powerful open-source desktop application that provides a 100% offline alternative to ChatGPT. It allows users to download and run various large language models locally, ensuring complete control and privacy over their AI interactions. With support for multiple platforms, Jan offers a robust solution for personal and private AI use.

TurboLLM: Run Local LLMs with Auto-Tuning, Any Engine, and a Polished UI
August 13, 2026
TurboLLM is a powerful, lightweight solution for running local LLM engines, offering auto-tuning for your GPU and a polished web UI. It supports any llama-server compatible binary, including community forks, and provides both OpenAI and Anthropic-compatible APIs. This offline-first tool allows users to maximize performance and flexibility with their local language models.
Source repository
Open the original repository on GitHub.
21 counted GitHub visits