Open-source coding agents compared

1. Aider

   Coding agent Standard

   Pair programs with an LLM in your terminal, editing files in your git repo and committing each change

   Pick it if you want every change recorded as a git commit and your choice of model provider, including local models.

   • 49.6k stars
   • 39.5 stars/day since creation
   • pushed May 22
   • v0.86.0 on Aug 9, 2025
   • Apache-2.0

   Install with pip and aider-install

   python -m pip install aider-install
   aider-install

   Copy install command for Aider
   Use it when, skip it if, stats for Aider

   Why you'd want it

   Aider builds a map of your whole codebase so it can work in larger projects, and it commits its changes to git with descriptive messages so you can diff and undo them with normal git tools. It connects to most cloud LLMs and to local models, can lint and test after every edit and fix what fails, and also takes requests from comments you add in your editor. Images, web pages and voice input can be added to the chat.

   Best for

   • Git-centric terminal workflows
   • Bring-your-own-model setups, including local models
   • Editing existing codebases

   Use it when

   You want an agent that records every change as a git commit and lets you choose the model provider, including local models.

   Skip it if

   You want an IDE-native agent with visual diffs, or you do not want to manage your own model API keys.

   Stats

   Stars

   49,554

   Last push

   May 22

   Release

   v0.86.0, Aug 9, 2025

   License

   Apache-2.0

   Pricing

   Open source

   Works with

   macOS, Linux, Windows

   Stars per day, since repo creation

   39.5

   Tier

   Standard, not ranked

   Full stat block for Aider

2. Gemini CLI

   Coding agent Standard

   Brings Gemini models into the terminal to query and edit codebases, run shell commands and search the web

   Pick it if your organization has a Gemini Code Assist Standard or Enterprise license, or pays for Gemini API or Vertex AI access.

   • 107k stars
   • 197 stars/day since creation
   • pushed Oct 14
   • v0.64.0 on Oct 13
   • Apache-2.0

   Install globally with npm

   npm install -g @google/gemini-cli

   Copy install command for Gemini CLI
   Use it when, skip it if, stats for Gemini CLI

   Why you'd want it

   Gemini CLI is a terminal agent with built-in file, shell, web fetch and Google Search grounding tools, and it adds MCP servers configured in ~/.gemini/settings.json. It reads project context from GEMINI.md, can checkpoint and resume sessions, and has a headless mode with JSON and streaming JSON output for scripts. A separate GitHub Action uses it for PR reviews and issue triage.

   Best for

   • Teams with a Gemini Code Assist Standard or Enterprise license
   • Google Cloud and Vertex AI users
   • Scripted runs that need JSON output

   Use it when

   Your organization has a Gemini Code Assist Standard or Enterprise license, or you pay for Gemini API or Vertex AI access, and you want a terminal agent on Gemini models.

   Skip it if

   You planned to use it for free with a personal Google account or through Google AI Pro or Ultra: Gemini CLI stopped serving those users on June 18, 2026 and Google points them to Antigravity CLI.

   Stats

   Stars

   107,372

   Last push

   Oct 14

   Release

   v0.64.0, Oct 13

   License

   Apache-2.0

   Pricing

   Open source

   Works with

   macOS, Linux, Windows, VS Code

   Stars per day, since repo creation

   197

   Tier

   Standard, not ranked

   Full stat block for Gemini CLI

3. OpenAI Codex CLI

   Coding agent Standard

   Runs OpenAI's coding agent locally to inspect your repo, edit files and run commands from the terminal

   Pick it if you have a ChatGPT plan or an OpenAI API key and want control over sandboxing and approvals on each run.

   • 129k stars
   • 235 stars/day since creation
   • pushed Oct 15
   • rust-v0.162.1 on Oct 9
   • Apache-2.0

   Install with the standalone installer (macOS, Linux)

   curl -fsSL https://chatgpt.com/codex/install.sh | sh

   Copy install command for OpenAI Codex CLI
   Use it when, skip it if, stats for OpenAI Codex CLI

   Why you'd want it

   Codex CLI works against your local repository with the tools already installed on your machine, and you choose the model, reasoning effort, permissions and sandbox for each run. You can use it interactively or call codex exec from scripts and CI, run a review against uncommitted changes, a commit or a base branch, add MCP servers with codex mcp, and hand work off to Codex cloud. You sign in with a ChatGPT plan or use an API key.

   Best for

   • ChatGPT Plus, Pro, Business, Edu or Enterprise users
   • Scripted runs in CI with codex exec
   • Reviewing changes before a commit

   Use it when

   You want a terminal coding agent on OpenAI models with control over sandboxing and approvals, and you have a ChatGPT plan or OpenAI API key.

   Skip it if

   You have neither a ChatGPT plan nor an OpenAI API key, or you would rather work in an editor, where OpenAI points you to its IDE extension or the Codex app.

   Stats

   Stars

   129,272

   Last push

   Oct 15

   Release

   rust-v0.162.1, Oct 9

   License

   Apache-2.0

   Pricing

   Open source

   Works with

   macOS, Linux, Windows, VS Code, Cursor, Windsurf

   Stars per day, since repo creation

   235

   Tier

   Standard, not ranked

   Full stat block for OpenAI Codex CLI

4. OpenCode

   Coding agent Standard

   Runs an open-source coding agent in your terminal, desktop app or browser with the LLM provider you choose

   Pick it if you want an MIT-licensed agent with a TUI, a desktop app and a local web interface that works with many model providers.

   • 214k stars
   • 401 stars/day since creation
   • pushed Oct 15
   • v1.19.0 on Oct 11
   • MIT

   Install OpenCode 2 with npm

   npm install -g @opencode/cli

   Copy install command for OpenCode
   Use it when, skip it if, stats for OpenCode

   Why you'd want it

   OpenCode has two built-in agents you switch between with Tab: build, with full access for development work, and plan, which is read-only, denies file edits by default and asks before running bash commands. It supports many LLM providers out of the box, connected from the TUI with /connect, and extends through MCP servers, plugins, agent skills and custom commands. The team also sells OpenCode Go, a subscription for open-source models, but you can use your own provider keys instead.

   Best for

   • Terminal users who want provider choice
   • Separating read-only planning from editing
   • Using the same agent from a TUI, desktop app or web UI

   Use it when

   You want an MIT-licensed coding agent that works with many model providers and offers a TUI, a desktop app and a local web interface.

   Skip it if

   You install CLI tools through Windows package managers, which OpenCode 2 does not support; also note OpenCode 1 and 2 cannot be installed side by side under the default command.

   Stats

   Stars

   213,573

   Last push

   Oct 15

   Release

   v1.19.0, Oct 11

   License

   MIT

   Pricing

   Open source

   Works with

   macOS, Linux, Windows, Web, Docker

   Stars per day, since repo creation

   401

   Tier

   Standard, not ranked

   Full stat block for OpenCode

5. Qwen Code

   Coding agent Standard

   Runs a coding agent in your terminal that works with Qwen, OpenAI, Anthropic, Gemini or local models

   Pick it if you want a terminal agent not tied to one model vendor, or Qwen models in particular. The free Qwen OAuth tier ended on 2026-04-15.

   • 28.5k stars
   • 59.8 stars/day since creation
   • pushed Oct 15
   • sdk-typescript-v0.1.19 on Oct 11
   • Apache-2.0

   Install with npm (Node.js 22+)

   npm install -g @qwen-code/qwen-code@latest

   Copy install command for Qwen Code
   Use it when, skip it if, stats for Qwen Code

   Why you'd want it

   Qwen Code was originally based on Gemini CLI v0.8.2 and has been developed independently since v0.1. It supports OpenAI, Anthropic, Gemini and Qwen APIs, other third-party providers and local models through Ollama or vLLM, and you can switch provider at runtime. Beyond the terminal it has VS Code, Zed and JetBrains integrations, a desktop app, an experimental web UI, a headless qwen -p mode and TypeScript, Python and Java SDKs.

   Best for

   • Running Qwen models in an agent
   • Mixing hosted and local models
   • Scripted runs in headless mode

   Use it when

   You want a terminal coding agent that is not tied to one model vendor, or you specifically want to use Qwen models.

   Skip it if

   You were counting on the free Qwen OAuth tier, which the docs say was discontinued on 2026-04-15; you now need an API key, an Alibaba Cloud plan or a local model. Its docs say optional usage statistics can be collected, so check the setting.

   Stats

   Stars

   28,487

   Last push

   Oct 15

   Release

   sdk-typescript-v0.1.19, Oct 11

   License

   Apache-2.0

   Pricing

   Open source

   Works with

   macOS, Linux, Windows, VS Code, JetBrains, Zed, Web

   Stars per day, since repo creation

   59.8

   Tier

   Standard, not ranked

   Full stat block for Qwen Code

1. Rank 15

   Kilo Code

   Coding agent

   Writes and edits code from natural language in VS Code, JetBrains or the terminal, using a model you pick

   Pick it if you want the same agent in VS Code, JetBrains and the terminal and want to switch models per task.

   • score 46
   • 27.6k stars
   • 47.3 stars/day since creation
   • pushed Oct 15
   • jetbrains/v7.1.9 on Oct 8
   • MIT

   Install the CLI with npm

   npm install -g @kilocode/cli

   Copy install command for Kilo Code
   Use it when, skip it if, stats for Kilo Code

   Why you'd want it

   Kilo Code ships Code, Plan, Ask and Debug agents, supports custom agents, and lets you switch between 500+ models mid-task at the provider's rate, per its README. It also offers inline autocomplete, terminal and browser control, and a marketplace for agents, skills, MCP servers and plugins. The CLI is a fork of OpenCode and has a kilo run --auto mode for CI pipelines.

   Best for

   • Teams split across VS Code and JetBrains
   • Switching models mid-task
   • Unattended agent runs in CI

   Use it when

   You want the same coding agent in VS Code, JetBrains and the terminal, and want to choose or switch the model per task.

   Skip it if

   The README's VS Code setup has you create a Kilo account to reach its model catalog, so skip it if you want no vendor account in the loop. If you already run OpenCode, the Kilo CLI adds little since it is a fork of it.

   Stats

   Stars

   27,607

   Last push

   Oct 15

   Release

   jetbrains/v7.1.9, Oct 8

   License

   MIT

   Pricing

   Open source

   Works with

   VS Code, JetBrains, macOS, Linux, Windows, Web

   Stars per day, since repo creation

   47.3

   New and rising score

   46 of 100

   Full stat block for Kilo Code

2. Cline

   Coding agent Standard

   Runs an approval-based coding agent in VS Code, JetBrains, the terminal or a desktop app with any model

   Pick it if you want an agent in VS Code or JetBrains that shows each edit as a diff and asks before each change.

   • 70.3k stars
   • 84.7 stars/day since creation
   • pushed Oct 15
   • desktop-v0.0.46 on Oct 13
   • Apache-2.0

   Install the CLI with npm

   npm i -g cline

   Copy install command for Cline
   Use it when, skip it if, stats for Cline

   Why you'd want it

   Cline edits files across a project, runs terminal commands and reacts to their output, and in VS Code and JetBrains each edit appears as a diff with checkpoints so you can undo its work. Plan mode explores the codebase and proposes a strategy before Act mode carries it out, and every edit and command needs your approval unless you turn on auto-approve. It works with Anthropic, OpenAI, Gemini, OpenRouter, Bedrock, Ollama, LM Studio and any OpenAI-compatible API, and extends through MCP servers, SDK plugins and .clinerules files.

   Best for

   • VS Code and JetBrains users
   • Developers who want to approve each action
   • Bring-your-own-model or local-model setups

   Use it when

   You want an agent inside your editor that asks before each change, and you want to pick the model provider yourself.

   Skip it if

   You need every client to be open source (the JetBrains plugin is not), or you want an agent that runs without per-step approvals by default.

   Stats

   Stars

   70,339

   Last push

   Oct 15

   Release

   desktop-v0.0.46, Oct 13

   License

   Apache-2.0

   Pricing

   Open source

   Works with

   VS Code, JetBrains, macOS, Windows

   Stars per day, since repo creation

   84.7

   Tier

   Standard, not ranked

   Full stat block for Cline

1. goose

   Coding agent Standard

   Runs a general-purpose AI agent on your machine as a desktop app, CLI or embeddable API

   Pick it if you want a local agent for code and other work, not tied to one vendor, that you extend with MCP servers.

   • 55.3k stars
   • 70.7 stars/day since creation
   • pushed Oct 15
   • v1.54.0 on Oct 8
   • Apache-2.0

   Install the CLI with the install script

   curl -fsSL https://github.com/aaif-goose/goose/releases/download/stable/download_cli.sh | bash

   Copy install command for goose
   Use it when, skip it if, stats for goose

   Why you'd want it

   goose is a local agent for code as well as research, writing, automation and data analysis, available as a desktop app for macOS, Linux and Windows, a CLI and an API. It works with 15+ providers including Anthropic, OpenAI, Google, Ollama, OpenRouter, Azure and Bedrock, and can use existing Claude, ChatGPT or Gemini subscriptions through ACP. It connects to 70+ extensions over MCP, and you can build custom distributions with preconfigured providers, extensions and branding.

   Best for

   • Agents for tasks beyond coding
   • Desktop app users who also want a CLI
   • Teams building an internal agent distribution

   Use it when

   You want an open-source local agent that is not tied to one model vendor and that you extend with MCP servers.

   Skip it if

   You want a coding agent built into your IDE; goose runs as its own desktop app or CLI.

   Stats

   Stars

   55,293

   Last push

   Oct 15

   Release

   v1.54.0, Oct 8

   License

   Apache-2.0

   Pricing

   Open source

   Works with

   macOS, Linux, Windows

   Stars per day, since repo creation

   70.7

   Tier

   Standard, not ranked

   Full stat block for goose

2. OpenHands

   Coding agent Standard

   Runs OpenHands, Claude Code, Codex and other ACP agents from a self-hosted control center

   Pick it if you want agents to keep working on a server or in Docker sandboxes and to start them from Slack or GitHub.

   • 90.9k stars
   • 96.1 stars/day since creation
   • pushed Oct 15
   • v1.27.1 on Oct 12
   • MIT

   Install Agent Canvas with npm

   npm install -g @openhands/agent-canvas
   agent-canvas

   Copy install command for OpenHands
   Use it when, skip it if, stats for OpenHands

   Why you'd want it

   The OpenHands repo now ships Agent Canvas, a self-hosted UI that runs the open-source OpenHands agent and can also drive Claude Code, Codex, Gemini CLI, OpenCode or any ACP-compatible agent. Agents run on Agent Server backends on your laptop, in Docker containers, on VMs or in OpenHands Cloud, and you switch between backends from one frontend. Automations run on a schedule or on webhook events and connect to Slack, GitHub, Linear and Notion.

   Best for

   • Teams running always-on agents on their own servers
   • Running several agents in isolated Docker containers
   • Scheduled and webhook-driven agent automations

   Use it when

   You want coding agents on a server or in Docker sandboxes so they keep working when your laptop is closed, and you want to trigger them from Slack or GitHub.

   Skip it if

   You just want one agent in your terminal or editor; the npm install runs the agent server directly on your machine with full filesystem access unless you use the Docker option, and it needs Node.js 24 and uv.

   Stats

   Stars

   90,928

   Last push

   Oct 15

   Release

   v1.27.1, Oct 12

   License

   MIT

   Pricing

   Open source

   Works with

   macOS, Linux, Windows, Docker

   Stars per day, since repo creation

   96.1

   Tier

   Standard, not ranked

   Full stat block for OpenHands

1. Rank 20

   Crush

   Coding agent

   Runs a terminal coding agent from Charm that works with your choice of LLM and uses LSPs for context

   Pick it if you want a terminal agent that uses your language servers and any model provider, and the FSL-1.1-MIT terms work for you.

   • score 42
   • 28.6k stars
   • 56 stars/day since creation
   • pushed Oct 15
   • v0.98.1 on Oct 9
   • FSL-1.1-MIT

   Install with npm

   npm install -g @charmland/crush

   Copy install command for Crush
   Use it when, skip it if, stats for Crush

   Why you'd want it

   Crush lets you pick from many LLM providers, or add your own through OpenAI- or Anthropic-compatible APIs and local runtimes such as Ollama and LM Studio, and switch models mid-session without losing context. It keeps multiple sessions per project, pulls extra context from language servers, and extends through MCP servers over http, stdio and sse, plus agent skills and hooks. It runs in terminals on macOS, Linux, Windows, Android, FreeBSD, OpenBSD and NetBSD.

   Best for

   • Terminal users on any OS, including the BSDs
   • Switching models mid-session
   • LSP-aware context for the agent

   Use it when

   You want a terminal agent that is not tied to one model vendor and can use the language servers you already run.

   Skip it if

   You need an OSI open-source license: Crush uses FSL-1.1-MIT, which bars competing commercial use until a version converts to MIT after two years.

   Stats

   Stars

   28,644

   Last push

   Oct 15

   Release

   v0.98.1, Oct 9

   License

   FSL-1.1-MIT

   Pricing

   Free

   Works with

   macOS, Linux, Windows

   Stars per day, since repo creation

   56

   New and rising score

   42 of 100

   Full stat block for Crush

1. Claude Code

   Coding agent Standard

   Reads your codebase, edits files, runs commands and handles git work from the terminal, IDE, desktop or web

   Not open source: Claude Code is proprietary, and most of its surfaces need a Claude subscription or an Anthropic Console account.

   • 151k stars
   • 252 stars/day since creation
   • pushed Oct 14
   • v2.1.300 on Oct 14
   • Proprietary

   Install with the native installer (macOS, Linux, WSL)

   curl -fsSL https://claude.ai/install.sh | bash

   Copy install command for Claude Code
   Use it when, skip it if, stats for Claude Code

   Why you'd want it

   Claude Code runs in the terminal, VS Code, JetBrains IDEs, a desktop app and the browser, and every surface shares the same CLAUDE.md files, settings and MCP servers. It plans multi-file changes, runs tests, stages commits and opens pull requests, and it can read an existing AGENTS.md. Subagents, skills, hooks and a non-interactive claude -p mode let you script it in CI, and the Agent SDK exposes the same tools for building your own agents.

   Best for

   • Developers who want one agent across terminal and IDE
   • Teams automating PR review and issue triage in CI
   • Builders working with Claude models

   Use it when

   You want an agent that works across a whole repo, runs your tools and handles git, and you have a Claude subscription or Anthropic Console account.

   Skip it if

   You need an open-source agent or have no Claude subscription or Anthropic Console account; Claude Code is proprietary and most surfaces require one of them.

   Stats

   Stars

   150,924

   Last push

   Oct 14

   Release

   v2.1.300, Oct 14

   License

   Proprietary

   Pricing

   Paid

   Works with

   VS Code, Cursor, JetBrains, macOS, Linux, Windows, Web

   Stars per day, since repo creation

   252

   Tier

   Standard, not ranked

   Full stat block for Claude Code

1. Rank 6

   Orca

   Devtool

   Runs several CLI coding agents side by side, each in its own git worktree, in one desktop app

   • score 61
   • 92.7k stars
   • 437 stars/day since creation
   • pushed Oct 15
   • v1.4.226 on Oct 15
   • MIT

   Download the app, or install with Homebrew on macOS

   brew install --cask stablyai/orca/orca

   Copy install command for Orca
   Use it when, skip it if, stats for Orca

   Why you'd want it

   Orca is a desktop app for running Claude Code, Codex, OpenCode, Pi and other terminal agents in parallel, each in an isolated git worktree, so you can fan one prompt across several agents, compare the results and merge the one you want. It combines terminal splits, VS Code's editor, an embedded Chromium with a Design Mode that sends an element's HTML, CSS and a screenshot into a prompt, GitHub and Linear task views, SSH worktrees on remote machines and inline comments on AI diffs. A CLI lets agents drive Orca themselves, and iOS and Android companion apps notify you when an agent finishes and let you send follow-ups.

   Best for

   • Comparing several agents on one task
   • Managing many worktrees in one place
   • Steering agents from a phone

   Use it when

   You run several CLI coding agents at once and want one window for their worktrees, terminals, diffs and reviews instead of juggling terminal tabs.

   Skip it if

   You prefer a terminal-only setup or use one agent at a time. It collects anonymous usage data unless you opt out, its Computer Use feature lets agents operate desktop apps, and it ships releases almost daily with a large open issue and pull request backlog.

   Stats

   Stars

   92,678

   Last push

   Oct 15

   Release

   v1.4.226, Oct 15

   License

   MIT

   Pricing

   Open source

   Works with

   Claude Code, Codex, Cursor, macOS, Windows, Linux

   Stars per day, since repo creation

   437

   New and rising score

   61 of 100

   Full stat block for Orca

2. Rank 7

   herdr

   Devtool

   Keeps coding agent terminals running in a background server and marks each pane as working, blocked or idle

   • score 59
   • 44.1k stars
   • 219 stars/day since creation
   • pushed Oct 14
   • v0.9.5 on Oct 14
   • Apache-2.0

   Install with Homebrew

   brew install herdr

   Copy install command for herdr
   Use it when, skip it if, stats for herdr

   Why you'd want it

   herdr is a terminal workspace for coding agents, a single Rust binary that runs inside the terminal you already use. Terminals keep running in a background server when you close the client or lose an SSH connection, and after a restart it restores the saved layout and can resume supported agent sessions. Every pane is marked working, blocked or idle, local and saved SSH machines share one window, and agents can drive herdr through its CLI and socket API to spawn panes, prompt each other and wait until another agent is blocked. It runs Claude Code, Codex, Cursor, OpenCode and other agents without wrapping them, with tmux-style prefix keys and mouse control.

   Best for

   • Running many terminal agents side by side
   • Detaching from long agent runs
   • Agents on remote machines over SSH

   Use it when

   You run several CLI coding agents at once, often over SSH, and want to detach, come back and see which one needs an answer.

   Skip it if

   You already have a tmux setup you like and run one agent at a time. Agents can use herdr's socket API to spawn panes and prompt other agents, and after a server or machine restart the original processes are gone, only the layout and supported sessions come back.

   Stats

   Stars

   44,074

   Last push

   Oct 14

   Release

   v0.9.5, Oct 14

   License

   Apache-2.0

   Pricing

   Open source

   Works with

   Claude Code, Codex, Cursor, macOS, Linux, Windows

   Stars per day, since repo creation

   219

   New and rising score

   59 of 100

   Full stat block for herdr

3. Rank 14

   nono

   Sandbox

   Runs coding agents in a least-privilege sandbox with per-tool filesystem, network and credential rules

   • score 47
   • 4.5k stars
   • 17.5 stars/day since creation
   • pushed Oct 14
   • v0.80.0 on Oct 13
   • Apache-2.0

   Install with Homebrew

   brew install nono

   Copy install command for nono
   Use it when, skip it if, stats for nono

   Why you'd want it

   nono wraps an agent such as Claude Code, Codex or Pi in a sandbox with no daemon, container or VM, and shows the exact filesystem, network and credential access before applying it. Profiles for common agents are pulled from the nono registry and can be extended into an editable JSON profile. Tools the agent calls, such as git, gh, curl or kubectl, can run in their own command sandboxes with separate filesystem and network policy, and credentials reach them through a proxy that can filter by HTTP method and path.

   Best for

   • Sandboxing Claude Code, Codex or Pi locally
   • Scoping credentials to one tool
   • Sharing reviewed agent profiles with a team

   Use it when

   You want to run a coding agent on your own machine with explicit filesystem, network and credential limits, without setting up containers or VMs.

   Skip it if

   You want VM or container isolation; the README says nono runs with no daemon, container or VM. APIs may still change before 1.0, and Windows is supported through WSL2. It runs agent commands and brokers credentials, so review a profile's grants before pulling it from the registry.

   Stats

   Stars

   4,504

   Last push

   Oct 14

   Release

   v0.80.0, Oct 13

   License

   Apache-2.0

   Pricing

   Open source

   Works with

   Claude Code, Codex, macOS, Linux, Windows

   Stars per day, since repo creation

   17.5

   New and rising score

   47 of 100

   Full stat block for nono

4. Rank 27

   CC Safety Net

   Devtool

   Blocks destructive git and file commands and reads of secret files before a coding agent's tool call runs

   • score 37
   • 1.6k stars
   • 5.6 stars/day since creation
   • pushed Oct 15
   • v2.6.12 on Oct 15
   • MIT

   Install into your coding CLIs with npx

   npx -y cc-safety-net@latest install
   npx -y cc-safety-net@latest doctor

   Copy install command for CC Safety Net
   Use it when, skip it if, stats for CC Safety Net

   Why you'd want it

   CC Safety Net installs as a pre-execution hook in coding CLIs including Claude Code, Codex, Cursor, Gemini CLI, OpenCode and Copilot CLI, and parses what each command does, so wrapping it in bash -c or python -c or reordering flags does not hide it. It blocks commands such as git reset --hard, git push --force and rm -rf on dangerous targets, and access to SSH keys, .env files, ~/.aws and coding CLI credentials. Presets are tuned in a local GUI, official rulebooks add blocks for Terraform, AWS, gcloud and Azure, and committing .cc-safety-net/ shares the policy through git.

   Best for

   • Stopping agent force pushes and hard resets
   • Keeping agents away from .env files and SSH keys
   • Teams running several coding CLIs

   Use it when

   You let a coding agent run shell commands and want a check that refuses destructive git and file operations before they execute.

   Skip it if

   You need containment: the README says it is not a sandbox and does not contain processes, set filesystem permissions or watch network egress. It is maintained mostly by one developer and needs Node.js 18 or higher.

   Stats

   Stars

   1,632

   Last push

   Oct 15

   Release

   v2.6.12, Oct 15

   License

   MIT

   Pricing

   Open source

   Works with

   Claude Code, Codex, Cursor, Gemini CLI, macOS, Linux, Windows

   Stars per day, since repo creation

   5.6

   New and rising score

   37 of 100

   Full stat block for CC Safety Net

More guides

• MCP servers for coding agents, picked by job
• Claude Code alternatives, by what you need instead
• AI agent frameworks compared
• Agent observability and eval tools
• Category: Coding agents