A terminal-first directory of AI coding agents (tools that can read/edit repos and run commands) plus the harnesses that orchestrate, sandbox, or extend them.
Standalone open-source projects with active communities.
-
Pi — Minimal, adaptable terminal coding harness from the pi-mono toolkit; unified LLM API, TUI, skills, and MCP support.
-
OpenCode — Terminal-native coding agent with 75+ provider support, LSP integration, and privacy-first design (formerly opencode-ai; now at opencode.ai).
-
Aider — Pair-programming agent for editing files via diffs/patches, with strong git and multi-file workflows.
-
Goose — Local, extensible agent that can execute, edit, and test; designed to run on-device and integrate with MCP.
-
Plandex — "Plan-first" CLI agent for building features across multiple files with structured steps and 2M token context.
-
Cline CLI — Model-agnostic autonomous agent for planning, file edits, command execution, and browser use.
-
Roo Code CLI — Multi-mode CLI agent (architect/code/debug/orchestrator modes); 22k+ stars, Claude-like terminal interface with skills and checkpoints.
-
Kilo Code CLI — Agentic engineering platform with CLI; orchestrator mode, 100s of LLMs, skills, and checkpointing (15k+ stars).
-
ForgeCode — AI pair programmer supporting 300+ models, with task management, custom agents, and large-scale refactor tooling.
-
OpenHands CLI — CLI entrypoint for the OpenHands agent (ex-OpenDevin lineage), focused on autonomous repo tasks.
-
Continue CLI — Open-source terminal extension for multi-model coding with local/privacy focus.
-
SWE-agent — Agent for resolving real repo issues/PR tasks; frequently used in SWE-bench-style workflows.
-
AutoCodeRover — Autonomous program improvement agent; patches real GitHub issues using code search and analysis.
-
Agentless — Lightweight approach to autonomous software engineering without persistent agent loops.
-
RA.Aid — Autonomous coding agent built on LangGraph with research/plan/implement pipeline; optional aider integration for near-full autonomy.
-
Devon — Open-source pair programmer with a TUI; autonomous planning, execution, and debugging in Git workflows.
-
Open Interpreter — Terminal tool that can execute code and actions; often used as a "do things on my machine" agent.
-
Claude Engineer — Community-driven CLI for agentic Claude workflows with file management and iterative development.
-
Crush — Charmbracelet's glamorous agentic coding TUI in Go; multi-provider, LSP-aware, with rich terminal UI.
-
gptme — Personal AI assistant in the terminal; runs code, manages files, and browses the web with tool-use.
-
g3 — "Coding AI agent" in Rust: tool-running, repo interaction, skills system, and provider abstraction.
-
Kode CLI — ShareAI's open-source CLI agent for terminal-native coding with multi-provider support.
-
OpenClaw — Personal AI assistant you run locally (170k+ stars); CLI with onboarding wizard, skills, tools, and multi-channel support (WhatsApp/Slack/Discord).
-
NanoClaw — Security-first lightweight alternative to OpenClaw; runs agents in Apple containers/Docker with sandboxed execution, built on Anthropic's Agents SDK.
-
PicoClaw — Ultra-lightweight personal AI assistant in Go inspired by OpenClaw; runs on $10 hardware with less than 10MB RAM.
-
IronClaw — OpenClaw rewritten in Rust by NEAR AI; WASM sandbox isolation, capability-based permissions, and prompt injection defense.
-
LettaBot — Personal AI assistant with persistent unified memory across Telegram, Slack, Discord, WhatsApp, and Signal; built on the Letta platform.
-
Dexto — Coding agent and general agent harness with CLI/web/API modes; ships a production-ready coding agent with sub-agent spawning.
-
2501 CLI — AI-powered autonomous CLI agent for coding, debugging production issues, and DevOps automation.
-
OH-MY-PI — Terminal coding agent ("Pi") with a TypeScript/Rust monorepo and local-first ergonomics.
-
Groq Code CLI — Customizable, lightweight CLI powered by Groq's ultra-fast inference; extensible tools/commands with multi-model support.
-
QQCode — Lightweight CLI coding agent in Rust focused on speed, determinism, and developer control; supports skills.
-
Coro Code — High-performance Rust CLI coding agent with rich terminal UI.
-
Codebuff — Multi-agent AI coding assistant with CLI support for collaborative coding workflows.
-
Codel — Autonomous agent for performing complex tasks via terminal; runs in Docker with a web UI.
-
cursor-agent — Python-based agent replicating Cursor's coding assistant capabilities; supports Claude, OpenAI, and local Ollama models.
-
Grok CLI — Community CLI agent built on xAI's Grok models for terminal-based coding tasks.
-
Every Code — A community "Codex-inspired" CLI agent focusing on local, scriptable coding workflows.
-
Mini-Kode — An educational AI coding agent CLI, intended as a readable reference implementation.
-
CodeMachine-CLI — Community multi-agent CLI aimed at running coding workflows locally (vibe-coding oriented).
-
CLAII — CLI-first AI coding agent with multi-agent orchestration, MCP toolchains, and memory-persistent refactors.
-
picocode — Minimal Rust-based coding agent focused on CI workflows and small codemods; multi-LLM with personas.
-
Binharic — A multi-provider "tech-priest persona" coding agent CLI (stylized, tool-using).
-
Smol Developer — Embeddable developer agent (12k+ stars) that generates entire codebases from a prompt; designed to be embedded in apps.
-
nori-cli — Multi-provider CLI built on Codex CLI; switch between Claude, Gemini, and Codex from the same native terminal.
-
open-codex — Lightweight fork of Codex CLI with multi-provider support (OpenAI, Gemini, OpenRouter, Ollama).
-
OpenDevin — Open-source agentic developer environment (not purely CLI, but often run locally with terminal entrypoints).
-
Letta Code — 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.
-
ZeroClaw — 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.
First-party CLI agents from major AI model providers and dev-tool companies.
-
Claude Code — Anthropic's repo-aware terminal agent for code edits, refactors, and git workflows.
-
Codex CLI — OpenAI's local coding agent for reading/editing/running code, with an interactive TUI and tool execution.
-
Gemini CLI — Google's open-source terminal agent powered by Gemini, with tools for repo work and research.
-
Amazon Q Developer CLI — AWS's agentic terminal chat for building apps, debugging, and DevOps with natural language.
-
Qwen Code — Alibaba Qwen's official CLI agent for Qwen coder models (workflow tool + repo operations).
-
Kimi CLI — Moonshot AI's open-source CLI coding agent with skills, MCP support, and ACP IDE integration.
-
Mistral Vibe — Mistral's open-source CLI coding assistant for conversational repo interaction and edits.
-
Trae Agent — ByteDance's research-friendly CLI agent for software engineering tasks, with modular architecture and multi-LLM support.
-
Neovate Code — Ant Group's open-source CLI agent with plugin system, multi-model/multi-provider support, MCP integrations, and headless automation mode.
-
FetchCoder — Terminal coding agent powered by ASI1, with interactive TUI, CLI, and API server modes plus MCP integration.
-
Amp — Sourcegraph's AI coding agent with a CLI for implementing tasks across real codebases.
-
Junie CLI — 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's data-native AI coding agent CLI for building pipelines, analytics, and AI apps with enterprise governance.
-
Devin — Cognition's autonomous AI software engineer with full shell/browser access, self-healing code, and PR collaboration.
-
GitHub Copilot in the CLI — GitHub's agentic CLI for repo/PR/issue workflows, command suggestions, and headless automation.
-
Cursor CLI — Cursor's official command-line agent (
agent) with shell mode, headless/CI support, parallel agents, and multi-model access. -
Mentat CLI — Cloud-native coding agent CLI for managing remote Mentat agents from your terminal; auto-detects repo/branch context.
-
Command Code — 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.
-
Catnip — Containerized environment + worktree automation for running multiple coding agents in parallel (optimized for Claude Code).
-
Claude Squad — tmux-based harness to run and manage multiple Claude Code sessions side-by-side.
-
Emdash — Run multiple coding agents concurrently with coordinated workflows.
-
Superset — A terminal built for coding agents; orchestrates parallel agent sessions.
-
Toad — Agent orchestrator for running and managing parallel CLI coding sessions.
-
amux — Terminal UI designed for running multiple coding agents in parallel.
-
cmux — Open-source platform for running multiple coding agents in parallel.
-
Crystal — Execute multiple Codex and Claude Code sessions in parallel git worktrees.
-
mux — Desktop application for isolated, parallel agentic development.
-
vibe-tree — Execute Claude Code tasks in parallel git worktrees.
-
Agent of Empires — tmux-based multi-agent harness for parallel terminal agents and task routing.
-
CLI Agent Orchestrator (CAO) — AWS's hierarchical multi-agent orchestration via tmux with intelligent task delegation patterns.
-
AgentPipe — Open-source, self-hosted platform for running and monitoring real-time multi-agent conversations across CLI tools.
-
multi-agent-workflow-kit — Orchestrate parallel AI agents in isolated git worktrees with shared tmux visibility.
-
vibe-kanban — Kanban interface for administering AI coding agents.
-
jean — Administer multiple projects, worktrees, and sessions with Claude CLI.
-
supacode — Native macOS coding agent orchestrator.
-
CodexMonitor — Coordinate multiple Codex agents across local workspaces.
Multi-agent coordination, swarm patterns, and autonomous execution loops.
-
claude-flow — Deploy multi-agent swarms with coordinated workflows.
-
gastown — Multi-agent orchestration with persistent work tracking.
-
loom — Infrastructure enabling autonomous loops to evolve products via multi-agent coordination.
-
ralph-orchestrator — Hat-based system maintaining agents in a loop until task completion.
-
ralph-tui — Direct AI agents through task lists with autonomous execution.
-
wreckit — Apply the Ralph Wiggum Loop pattern across your roadmap for autonomous agent execution.
Sandboxes, routers, browser/terminal automation, and extension tools.
-
AgentManager — Lightweight CLI for managing multiple agent runs/sessions and workflows.
-
AgentDeck — CLI-driven "deck" of agents with configs/presets for different tasks and repos.
-
Claude Code Tools — Utilities around Claude Code workflows (automation helpers, helpers for common tasks).
-
claude-cmd — Terminal wrapper for interacting with Claude models; often used as a building block in harness scripts.
-
claude-code-router — Route Claude Code calls to alternative providers/endpoints (useful for self-hosting proxies).
-
claudebox — Sandboxed environment for Claude Code (focused on isolation/safety).
-
agent-browser — Headless browser automation CLI for agents (useful as a tool plugin for coding agents).
-
agent-terminal — Headless terminal automation for AI agents using node-pty; capture output and send input programmatically.
-
subtask — Claude Skill for delegating tasks with subagents in Git worktrees.
-
claude-northstar — Transforms CLI agents from task executors into autonomous project partners.
-
open-claude-cowork — Open-source version of Claude Cowork with 500+ SaaS app integrations.
-
pi-builder — 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.
PRs welcome! Please keep each entry:
- Name + link
- 1–2 line description (what it does, who it's for)
Optional: add a "Why it's interesting" note (diffs, LSP, sandboxing, multi-agent, MCP/skills, etc).
