Awesome CLI Coding Agents

A curated list of 80+ CLI coding agents — AI-powered tools that live in your terminal, read/edit repos, and run commands — plus the harnesses that orchestrate, sandbox, or extend them.

Last updated: 2026-04-06

A CLI coding agent is an AI-powered tool that runs in your terminal and can autonomously read, write, and execute code in your repository. Unlike chat-based assistants, these agents have direct access to your filesystem, shell, and dev tools — they can edit files, run tests, commit changes, and iterate on errors. Think of them as AI pair programmers that live where you already work: the command line.

• Terminal-native coding agents

  • Open Source
  • OpenClaw ecosystem
  • Closed Source

• Harnesses & orchestration

  • Session managers & parallel runners
  • Orchestrators & autonomous loops
  • Agent infrastructure

• Contributing

Forkable, extensible, and community-driven. Sorted by GitHub stars. Provider tags [Company] indicate the backing organization.

• OpenCode ⭐ 122k — Terminal-native coding agent with 75+ provider support, LSP integration, and privacy-first design (formerly opencode-ai; now at opencode.ai).

• Claw Code ⭐ 110k — Clean-room Python/Rust rewrite of Claude Code architecture using oh-my-codex; fastest repo in GitHub history to 100K stars. Born from the March 2026 Claude Code source leak. MIT.

• Gemini CLI ⭐ 98k [Google] — Google's terminal agent powered by Gemini, with tools for repo work and research. Apache-2.0.

• OpenHands ⭐ 69.3k — Open-source agentic developer environment (formerly OpenDevin) with CLI and web entrypoints; also has a lightweight CLI-only package.

• Codex CLI ⭐ 66.1k [OpenAI] — OpenAI's local coding agent for reading/editing/running code, with an interactive TUI and tool execution. Apache-2.0.

• Open Interpreter ⭐ 63k — Terminal tool that can execute code and actions; often used as a "do things on my machine" agent.

• Cline CLI ⭐ 59k — Model-agnostic autonomous agent for planning, file edits, command execution, and browser use.

• Aider ⭐ 42k — Pair-programming agent for editing files via diffs/patches, with strong git and multi-file workflows.

• Goose ⭐ 33k — Local, extensible agent that can execute, edit, and test; designed to run on-device and integrate with MCP.

• Continue CLI ⭐ 32k — Open-source terminal extension for multi-model coding with local/privacy focus.

• Pi ⭐ 25.4k — Minimal, adaptable terminal coding harness from the pi-mono toolkit; unified LLM API, TUI, skills, and MCP support.

• Roo Code CLI ⭐ 22.7k — Multi-mode CLI agent (architect/code/debug/orchestrator modes); Claude-like terminal interface with skills and checkpoints.

• Crush ⭐ 21.6k — Charmbracelet's glamorous agentic coding TUI in Go; multi-provider, LSP-aware, with rich terminal UI.

• Qwen Code ⭐ 20k [Alibaba] — Alibaba Qwen's official CLI agent for Qwen coder models (workflow tool + repo operations). Apache-2.0.

• SWE-agent ⭐ 18.8k — Agent for resolving real repo issues/PR tasks; frequently used in SWE-bench-style workflows.

• Kilo Code CLI ⭐ 16.9k — Agentic engineering platform with CLI; orchestrator mode, 100s of LLMs, skills, and checkpointing.

• Plandex ⭐ 15k — "Plan-first" CLI agent for building features across multiple files with structured steps and 2M token context.

• Smol Developer ⭐ 12k — Embeddable developer agent that generates entire codebases from a prompt; designed to be embedded in apps.

• Claude Engineer ⭐ 11k — Community-driven CLI for agentic Claude workflows with file management and iterative development.

• Trae Agent ⭐ 11k [ByteDance] — ByteDance's research-friendly CLI agent for software engineering tasks, with modular architecture and multi-LLM support. MIT.

• Hermes Agent ⭐ 8.7k — Nous Research's self-improving CLI agent with persistent memory, automated skill creation, sandboxed code execution via Unix socket RPC, and multi-platform reach (Telegram/Slack/Discord/WhatsApp); supports 300+ models across multiple providers.

• Kimi CLI ⭐ 7.1k [Moonshot AI] — Moonshot AI's CLI coding agent with skills, MCP support, and ACP IDE integration. Apache-2.0.

• Claurst ⭐ 6.1k — Claude Code rewritten in idiomatic Rust with architectural breakdown; includes discoveries from the source leak (KAIROS persistent assistant, buddy system). MIT.

• ForgeCode ⭐ 5.0k — AI pair programmer supporting 300+ models, with task management, custom agents, and large-scale refactor tooling.

• Kode CLI ⭐ 4.6k — ShareAI's open-source CLI agent for terminal-native coding with multi-provider support.

• gptme ⭐ 4.2k — Personal AI assistant in the terminal; runs code, manages files, and browses the web with tool-use.

• Codebuff ⭐ 4.2k — Multi-agent AI coding assistant with CLI support for collaborative coding workflows.

• Every Code ⭐ 3.6k — Fork of Codex CLI with validation, automation, browser integration, multi-agents, theming, and multi-provider orchestration (OpenAI, Claude, Gemini). Apache-2.0.

• Devon ⭐ 3.5k — Open-source pair programmer with a TUI; autonomous planning, execution, and debugging in Git workflows.

• Mistral Vibe ⭐ 3.4k [Mistral] — Mistral's CLI coding assistant for conversational repo interaction and edits. Apache-2.0.

• Free Code ⭐ 3.3k — Fork of Claude Code with all telemetry removed, guardrails stripped, and all experimental features enabled (KAIROS, dream mode, companion system).

• AutoCodeRover ⭐ 3.1k — Autonomous program improvement agent; patches real GitHub issues using code search and analysis.

• Codel ⭐ 2.4k — Autonomous agent for performing complex tasks via terminal; runs in Docker with a web UI.

• Grok CLI ⭐ 2.4k — Community CLI agent built on xAI's Grok models for terminal-based coding tasks.

• CodeMachine-CLI ⭐ 2.4k — Community multi-agent CLI aimed at running coding workflows locally (vibe-coding oriented).

• RA.Aid ⭐ 2.2k — Autonomous coding agent built on LangGraph with research/plan/implement pipeline; optional aider integration for near-full autonomy.

• open-codex ⭐ 2.2k — Lightweight fork of Codex CLI with multi-provider support (OpenAI, Gemini, OpenRouter, Ollama).

• Agentless ⭐ 2.0k — Lightweight approach to autonomous software engineering without persistent agent loops.

• Amazon Q Developer CLI ⭐ 1.9k [AWS] — AWS's agentic terminal chat for building apps, debugging, and DevOps with natural language. Apache-2.0.

• OH-MY-PI ⭐ 2.1k — Terminal coding agent ("Pi") with a TypeScript/Rust monorepo and local-first ergonomics.

• Letta Code ⭐ 1.9k — Memory-first CLI coding agent built on the Letta platform (formerly MemGPT); persistent memory across sessions, model-agnostic (Claude/GPT/Gemini), skill learning, and context repositories.

• Neovate Code ⭐ 1.5k [Ant Group] — Ant Group's CLI agent with plugin system, multi-model/multi-provider support, MCP integrations, and headless automation mode. MIT.

• Groq Code CLI ⭐ 710 — Customizable, lightweight CLI powered by Groq's ultra-fast inference; extensible tools/commands with multi-model support.

• Dexto ⭐ 596 — Coding agent and general agent harness with CLI/web/API modes; ships a production-ready coding agent with sub-agent spawning.

• g3 ⭐ 477 — "Coding AI agent" in Rust: tool-running, repo interaction, skills system, and provider abstraction.

• Coro Code ⭐ 350 — Open-source CLI coding agent, a free alternative to Claude Code; generate, debug, and manage code seamlessly.

• Mini-Kode ⭐ 292 — An educational AI coding agent CLI, intended as a readable reference implementation.

• LettaBot ⭐ 267 — Personal AI assistant with persistent unified memory across Telegram, Slack, Discord, WhatsApp, and Signal; built on the Letta platform.

• cursor-agent ⭐ 112 — Python-based agent replicating Cursor's coding assistant capabilities; supports Claude, OpenAI, and local Ollama models.

• nori-cli ⭐ 113 — Multi-provider CLI built on Codex CLI; switch between Claude, Gemini, and Codex from the same native terminal.

• QQCode ⭐ 49 — Lightweight CLI coding agent in Rust focused on speed, determinism, and developer control; supports skills.

• picocode ⭐ 38 — Minimal Rust-based coding agent focused on CI workflows and small codemods; multi-LLM with personas.

• Binharic ⭐ 15 — A multi-provider "tech-priest persona" coding agent CLI (stylized, tool-using).

• CLAII — CLI-first AI coding agent with multi-agent orchestration, MCP toolchains, and memory-persistent refactors.

Projects built on, forked from, or inspired by OpenClaw — the open-source personal AI assistant. Sorted by GitHub stars.

• OpenClaw ⭐ 322k — The original personal AI assistant you run locally; CLI with onboarding wizard, skills, tools, and multi-channel support (WhatsApp/Slack/Discord). MIT.

• nanobot ⭐ 34.6k — Ultra-lightweight ~4,000-line Python rewrite of OpenClaw; tool use, persistent memory, scheduled tasks, and multi-channel support (Telegram/Discord/WhatsApp). MIT.

• ZeroClaw ⭐ 27.8k — Fully autonomous AI agent runtime in Rust; trait-driven pluggable architecture (providers, tools, memory, channels), runs on minimal hardware (<5MB RAM), multi-channel CLI/Telegram/Discord/Slack, with sandboxed execution and hybrid vector+keyword search.

• PicoClaw ⭐ 25.3k — Ultra-lightweight personal AI assistant in Go inspired by OpenClaw; runs on $10 hardware with less than 10MB RAM.

• NanoClaw ⭐ 24.0k — Security-first lightweight alternative to OpenClaw; runs agents in Apple containers/Docker with sandboxed execution, built on Anthropic's Agents SDK.

• IronClaw ⭐ 10.4k — OpenClaw rewritten in Rust by NEAR AI; WASM sandbox isolation, capability-based permissions, and prompt injection defense.

• NullClaw ⭐ 6.5k — Fastest, smallest OpenClaw-compatible agent in Zig; 678KB static binary, ~1MB RAM, <2ms startup, 23+ providers, 18 channels. MIT.

• Moltis ⭐ 2.3k — Secure, auditable Rust-native alternative to OpenClaw; zero unsafe code, 2,300+ tests, built-in voice I/O, MCP servers, and multi-channel access. MIT.

• Clawith ⭐ 1.6k — "OpenClaw for Teams" — multi-agent collaboration platform that scales OpenClaw to organizations. Apache-2.0.

• GitClaw ⭐ 149 — Git-native AI agent framework where agent identity, rules, memory, tools, and skills are all version-controlled files. MIT.

Proprietary agents — usable but not forkable or extensible at the source level.

• Claude Code ⭐ 79.5k [Anthropic] — Anthropic's repo-aware terminal agent for code edits, refactors, and git workflows. Source-available, no OSS license.

• Warp ⭐ 26.2k [Warp] — Modern terminal with built-in AI agent mode; understands tasks, runs commands, edits files, and orchestrates multi-step workflows.

• GitHub Copilot in the CLI ⭐ 9.4k [GitHub] — GitHub's agentic CLI for repo/PR/issue workflows, command suggestions, and headless automation.

• Droid ⭐ 624 [Factory] — Factory's multi-model CLI coding agent; #1 on Terminal-Bench, specialized droids for different tasks, headless CI mode, and multi-interface support (CLI/IDE/Slack/Linear).

• Amp [Sourcegraph] — Sourcegraph's AI coding agent with a CLI for implementing tasks across real codebases.

• Junie CLI [JetBrains] — JetBrains' LLM-agnostic CLI coding agent (EAP); supports GPT-5, Claude, Gemini, Grok with plan mode and CI/CD headless usage.

• Cortex Code CLI [Snowflake] — Snowflake's data-native AI coding agent CLI for building pipelines, analytics, and AI apps with enterprise governance.

• Devin [Cognition] — Cognition's autonomous AI software engineer with full shell/browser access, self-healing code, and PR collaboration.

• Cursor CLI [Cursor] — Cursor's official command-line agent (agent) with shell mode, headless/CI support, parallel agents, and multi-model access.

• Tabnine CLI [Tabnine] — AI-powered terminal coding assistant with agentic workflows; distributed as a Docker container, requires enterprise license.

• Mentat CLI [Mentat] — Cloud-native coding agent CLI for managing remote Mentat agents from your terminal; auto-detects repo/branch context.

• FetchCoder [Fetch.ai] — Terminal coding agent powered by ASI1, with interactive TUI, CLI, and API server modes plus MCP integration.

• Command Code [CommandCode] — CLI coding agent that continuously learns your coding style via taste-1 neuro-symbolic AI; adapts to preferences over time with project-specific taste profiles.

Tools for running and managing multiple agent sessions side-by-side. Sorted by GitHub stars.

• vibe-kanban ⭐ 23.4k — Kanban interface for administering AI coding agents.

• cmux ⭐ 8.1k — Open-source platform for running multiple coding agents in parallel.

• Superset ⭐ 7.4k — A terminal built for coding agents; orchestrates parallel agent sessions.

• Claude Squad ⭐ 6.4k — tmux-based harness to run and manage multiple Claude Code sessions side-by-side.

• CodexMonitor ⭐ 3.2k — Coordinate multiple Codex agents across local workspaces.

• Crystal ⭐ 3.0k — Execute multiple Codex and Claude Code sessions in parallel git worktrees.

• Toad ⭐ 2.7k — Agent orchestrator for running and managing parallel CLI coding sessions.

• Emdash ⭐ 2.7k — Run multiple coding agents concurrently with coordinated workflows.

• mux ⭐ 1.4k — Desktop application for isolated, parallel agentic development.

• jean ⭐ 658 — Administer multiple projects, worktrees, and sessions with Claude CLI.

• supacode ⭐ 511 — Native macOS coding agent orchestrator.

• Catnip ⭐ 467 — Containerized environment + worktree automation for running multiple coding agents in parallel (optimized for Claude Code).

• Parallel Code ⭐ 385 — Desktop app for running multiple AI coding agents (Claude Code, Codex CLI, Gemini CLI) simultaneously in isolated git worktrees.

• CLI Agent Orchestrator (CAO) ⭐ 330 — AWS's hierarchical multi-agent orchestration via tmux with intelligent task delegation patterns.

• vibe-tree ⭐ 244 — Execute Claude Code tasks in parallel git worktrees.

• AgentPipe ⭐ 98 — CLI/TUI app that orchestrates multi-agent conversations by enabling different AI CLI tools (Claude Code, Gemini, Qwen, etc.) to communicate in shared rooms. MIT.

• amux ⭐ 56 — Terminal UI designed for running multiple coding agents in parallel.

• CliDeck ⭐ 35 — WhatsApp-like browser dashboard for managing multiple CLI coding agents (Claude Code, Codex, Gemini CLI, OpenCode) with live status detection, session resume, autopilot routing, and full control from a phone while away. MIT.

• multi-agent-workflow-kit ⭐ 9 — Orchestrate parallel AI agents in isolated git worktrees with shared tmux visibility.

• PATAPIM — Terminal IDE with a 9-terminal grid for running multiple CLI coding agents simultaneously; features AI state detection, built-in Whisper voice dictation, LAN remote access, and an embedded MCP browser. Built with Electron and node-pty. Freemium.

Multi-agent coordination, swarm patterns, and autonomous execution loops. Sorted by GitHub stars.

• claude-flow ⭐ 21.6k — Deploy multi-agent swarms with coordinated workflows.

• gastown ⭐ 12.5k — Multi-agent orchestration with persistent work tracking.

• ralph-orchestrator ⭐ 2.2k — Hat-based system maintaining agents in a loop until task completion.

• ralph-tui ⭐ 2.1k — Direct AI agents through task lists with autonomous execution.

• loom ⭐ 1.2k — Infrastructure enabling autonomous loops to evolve products via multi-agent coordination.

• wreckit ⭐ 121 — Apply the Ralph Wiggum Loop pattern across your roadmap for autonomous agent execution.

• kodo ⭐ 37 — Autonomous multi-agent coding orchestrator that directs Claude Code, Codex, and Gemini CLI through work cycles with independent architect and tester verification. SWE-bench verified.

• OpenCastle ⭐ 18 — Multi-agent orchestration framework that turns AI coding assistants (Copilot, Cursor, Claude Code, OpenCode, Windsurf, Codex CLI) into 19 coordinated specialist agents. CLI-driven (npx opencastle init), with task decomposition, parallel work, and quality gates. MIT.

• Bernstein — Deterministic Python orchestrator — spawns parallel AI coding agents (Claude Code, Codex CLI, Gemini CLI), verifies with tests, auto-commits.

• ORCH ⭐ 9 — CLI orchestrator that manages Claude Code, Codex, and Cursor as a typed task queue with state machine (todo→in_progress→review→done), auto-retry, inter-agent messaging, and TUI dashboard.

• Forge ⭐ 4 — Autonomous spec-driven development loop for Claude Code; three-command pipeline (brainstorm specs, plan task DAGs, execute autonomously) with context survival, backpropagation that traces bugs to spec gaps, and Claude-on-Claude code review. MIT.

Sandboxes, routers, browser/terminal automation, and extension tools. Sorted by GitHub stars.

• claude-code-router ⭐ 29.9k — Use Claude Code as a foundation while routing to alternative providers/endpoints.

• agent-browser ⭐ 23.3k — Headless browser automation CLI for agents (useful as a tool plugin for coding agents).

• OpenWork ⭐ 13.0k — Open-source alternative to Claude Cowork for teams; local-first desktop app powered by OpenCode with one-click setup. MIT.

• NemoClaw ⭐ 8.0k [NVIDIA] — CLI tool for securely provisioning and managing sandboxed OpenClaw agent environments; enforces network, filesystem, and process-level security policies via OpenShell runtime. Apache-2.0.

• open-claude-cowork ⭐ 3.2k — Open-source version of Claude Cowork with 500+ SaaS app integrations.

• Claude Code Tools ⭐ 1.6k — Utilities around Claude Code workflows (automation helpers, helpers for common tasks).

• OneCLI ⭐ 809 — Open-source credential vault for AI agents; Rust HTTP gateway injects API keys transparently so agents never handle raw secrets. Per-agent scoped tokens, AES-256-GCM encryption at rest. Apache-2.0.

• subtask ⭐ 320 — Claude Skill for delegating tasks with subagents in Git worktrees.

• claude-cmd ⭐ 294 — Terminal wrapper for interacting with Claude models; often used as a building block in harness scripts.

• Coasts ⭐ 59 — Containerized host orchestration for git worktrees; isolates ports, networks, and services per environment using DinD/Sysbox/Podman. MIT.

• claudebox ⭐ 36 — Sandboxed environment for Claude Code (focused on isolation/safety).

• AgentManager ⭐ 17 — Lightweight CLI for managing multiple agent runs/sessions and workflows.

• brood-box ⭐ 11 — Hardware-isolated microVM sandbox for AI coding agents (Claude Code, Codex, OpenCode) with COW snapshot isolation, egress control, and MCP authorization.

• agent-terminal ⭐ 10 — Headless terminal automation for AI agents using node-pty; capture output and send input programmatically.

• Untether ⭐ 8 — Telegram bridge for 6 CLI coding agents (Claude Code, Codex, OpenCode, Pi, Gemini CLI, Amp); remote task control via voice or text, progress streaming, interactive permissions, and cost tracking. MIT.

• pi-builder ⭐ 5 — TypeScript monorepo that wraps any installed CLI coding agent (Claude Code, Aider, OpenCode, Codex, Gemini CLI, Goose, Plandex, SWE-agent, Crush, gptme) behind a single interface; capability-based routing, health caching, fallback chains, SQLite persistence, and a streaming OrchestratorService. MIT.

• zosma-qa ⭐ 5 — Generates QA agent prompts (planner, generator, healer, analyzer) for CLI coding tools (OpenCode, Claude Code, VS Code Copilot); scaffolds autonomous test workflows across Playwright, Appium, and k6.

• Wit ⭐ 4 — Coordination protocol that prevents merge conflicts between parallel AI agents. Locks specific functions (not files) via Tree-sitter AST parsing; agents declare intents, acquire symbol-level locks, and get conflict warnings before writing code. JSON-RPC daemon. MIT.

• OSOP ⭐ 0 — Universal workflow logging protocol for CLI coding agents; produces .osop workflow definitions and .osoplog.yaml execution records. Supports Claude Code, Codex, Cursor, Windsurf, Aider, Cline, Roo Code, Devin, and OpenClaw. Includes a visual editor and spec.

• Agentic Engineering Framework — Provider-neutral governance framework for CLI coding agents. Structural enforcement of task-driven workflows, context budget management, antifragile healing loops, and audit compliance. Works with Claude Code, Aider, Cursor, and any file-based agent.

• claude-northstar — Transforms CLI agents from task executors into autonomous project partners.

• kasetto — A declarative AI agent environment manager, written in Rust.

PRs welcome! To add an entry, please ensure it meets these criteria:

Inclusion requirements:

• Must have a CLI or terminal interface (IDE-only tools don't qualify)
• Must be able to read/write code or run commands autonomously
• Link must point to a valid, active project (no dead repos)

Entry format:

1. Name + link (GitHub preferred)
2. Star count (for GitHub repos)
3. 1–2 line description — what it does, who it's for

Optional: provider tag [Company], license, or a "why it's interesting" note (diffs, LSP, sandboxing, multi-agent, MCP/skills, etc).

Entries are sorted by GitHub stars within each section. Place your entry in the correct position.