Preview. These maps are our reading of public sources and haven't been checked with the projects' maintainers yet.
Coding agents
AI agents that read, write, and run code in a developer's repo.
22 projects68% active95% build on a shared standardMapped 2026-10-25
What they agree on
- Is it actively maintained? active (73% of 22)
- How is it mainly sustained? company with a paid product, tier, or ads (73% of 22)
- Who is it for? developers (95% of 22)
- Where do the commands it runs execute by default? on your machine with your user's permissions (86% of 22)
- Can users add their own tools through MCP (Model Context Protocol) servers? yes (95% of 22)
What they argue about
How is the main thing licensed?
Can it run entirely on your own device or infrastructure, without the maker's servers?
Do you need an account to use it?
Which shared standard or format does it mainly build on?
Where do you primarily use it?
How does finished work usually come back to you?
Whose models can you use? Pick the first that applies.
Is pointing it at a model running on your own machine (e.g. Ollama, LM Studio) a documented setup?
Out of the box, does it ask you before editing files or running shell commands?
Does it read project instructions from AGENTS.md?
Does it use the Agent Client Protocol (ACP) for agent-editor connections?
When you first set it up, how do you pay for the models it uses by default?
Projects
| Project | Status | What it is |
|---|---|---|
| Aider | Maintenance | Terminal AI pair programmer in Python that edits files in your git repo and auto-commits, works with almost any LLM; one of the earliest op… |
| Amp | Active | Proprietary coding agent from the Sourcegraph spin-out: CLI, web and mobile apps, and 'orbs' (vendor-run cloud machines) for background wor… |
| Claude Code | Active | Anthropic's agentic coding tool: terminal CLI first, plus VS Code/JetBrains extensions, a desktop app, and cloud sessions on the web that r… |
| Cline | Active | Open-source coding agent originally a VS Code extension, now also CLI, desktop app, JetBrains plugin and SDK; human-in-the-loop approvals,… |
| Continue | Acquired | Open-source coding assistant/agent for VS Code, JetBrains and CLI (cn) with configurable models and rules; acqui-hired by Cursor in June 20… |
| Crush | Active | Charm's Go terminal coding agent (glamorous TUI) with LSP context, MCP, multi-provider via the Catwalk model registry, and an optional Char… |
| Cursor | Acquired | VS Code-fork AI editor from Anysphere, now owned by SpaceX (with xAI). Agent in the editor, a CLI, and cloud agents; plans now bundle a usa… |
| Devin | Active | Cognition's autonomous cloud software engineer: tasks from Slack/Linear/GitHub/web run in a Devin VM and come back as PRs. Now also a Devin… |
| Factory Droid | Active | Factory's proprietary 'Droid' agent: terminal CLI plus desktop Factory App, headless droid exec, delegations from Slack/Linear/Jira, cloud… |
| Gemini CLI | Maintenance | Google's open-source (Apache-2.0) terminal agent for Gemini models, an early ACP agent. Now superseded by Antigravity CLI for most users; t… |
| GitHub Copilot | Active | GitHub/Microsoft's assistant, mapped as one product: agent mode in VS Code and other IDEs, Copilot CLI, and a cloud coding agent that works… |
| Goose | Active | General-purpose open-source agent (desktop app, CLI, API) started by Block, built around MCP extensions; now governed by the Linux Foundati… |
| Kilo Code | Active | Open-source coding agent for VS Code, JetBrains and CLI (began as a Roo/Cline fork, now built on OpenCode) with a zero-markup model gateway… |
| Kiro | Active | AWS's spec-driven agentic IDE (VS Code-based), now one agent harness across IDE, CLI, web (PR-opening background agent), mobile and 'Crew';… |
| OpenAI Codex CLI | Active | OpenAI's open-source (Apache-2.0) Rust coding agent for the terminal, sharing an engine with the Codex IDE extension, desktop app and Codex… |
| OpenCode | Active | Open-source terminal coding agent (also desktop app and IDE extension) from Anomaly, the SST team; 75+ providers via Models.dev, with an op… |
| OpenHands | Active | MIT-licensed open agent platform (formerly OpenDevin) from All Hands AI: Python Software Agent SDK and agent server, the OpenHands CLI, and… |
| Qwen Code | Active | Alibaba Qwen team's terminal coding agent, forked from Gemini CLI v0.8.2 and developed independently since; multi-protocol (OpenAI/Anthropi… |
| Roo Code | Changed direction | Popular Cline fork VS Code extension with custom modes (Architect, Code, Debug, Orchestrator); shut down in May 2026 as the team pivoted to… |
| Void | Quiet | Open-source VS Code fork AI editor (YC-backed) with agent mode, any-provider and local-model support; team paused it to explore new ideas a… |
| Windsurf | Changed direction | Former Codeium/Windsurf AI editor, now Cognition's Devin Desktop: a VS Code-based IDE with an Agent Command Center that runs Devin Local, c… |
| Zed | Active | Open-source Rust code editor with a built-in agent panel (Zed Agent) and host for external agents (Claude Agent, Codex, Gemini CLI, etc.) v… |
Where they meet
Not mapped yet.
Follow
RSS for this community: projects added, status changes, and later polls and discussions.
Follow the whole community: OPML of every project's own release feed (where we found one) · OPML of every project's Poll Position feed.
The poll
Every project above was answered against these questions. Core questions are shared by all communities; the rest are this community's own.
- Is it actively maintained? core
active · active, but changed direction (pivot, fork, or rename) · maintenance or slowing · dormant, sunsetting, or archived - How is the main thing licensed? core
permissive open source (MIT/Apache/BSD) · copyleft or source-available with restrictions · open parts around a closed core · proprietary or closed source - How is it mainly sustained? core
individual side project, no revenue · individual with revenue (paid app, tier, ads, or donations) · community, foundation, or research · company with a paid product, tier, or ads · company or startup with a free product (no revenue yet) · big company side project - Can it run entirely on your own device or infrastructure, without the maker's servers? core
yes · yes, but it calls a third-party service you choose (your own API key or account) · partly (some features need the maker's servers) · no, it needs the maker's servers · not applicable - Do you need an account to use it? core
no account · optional account · account required - Who is it for? core
developers · non-developers · both - Is there public evidence it's built with AI coding agents? core
yes, the repo shows it (agent instruction files or agent co-authored commits) · yes, the maker says so · no public evidence - Which shared standard or format does it mainly build on? core
MCP (Model Context Protocol) · AGENTS.md · ACP (Agent Client Protocol) · none of these (its own format or interface) · not applicable - Where do you primarily use it? approach
terminal (CLI or TUI) · extension inside an existing IDE · its own editor or desktop app · in a browser (a local web UI, a web app, or a cloud service) - How does finished work usually come back to you? approach
edits in your local working tree while you watch · a branch or pull request from a background run · both are first-class - Whose models can you use? Pick the first that applies. value
any provider or endpoint you configure (your own key, custom URL, or local model) · several vendors' models, only as served by the tool · one model vendor's models only - Is pointing it at a model running on your own machine (e.g. Ollama, LM Studio) a documented setup? approach
yes, documented · no - Out of the box, does it ask you before editing files or running shell commands? value
asks before edits and before commands · edits without asking, asks before most commands · doesn't ask (a sandbox or automatic reviewer may still block risky actions) - Where do the commands it runs execute by default? value
on your machine with your user's permissions · on your machine inside an OS sandbox or container · in a remote VM or container run by the vendor - Can users add their own tools through MCP (Model Context Protocol) servers? approach
yes · no - Does it read project instructions from AGENTS.md? value
yes, AGENTS.md is its main instructions file · yes, alongside its own file · no - Does it use the Agent Client Protocol (ACP) for agent-editor connections? approach
yes, as an agent that editors can drive · yes, as an editor that hosts ACP agents · both · no - When you first set it up, how do you pay for the models it uses by default? approach
with your own API key or local model (the tool itself is free) · a subscription plan from the tool's maker · pay-as-you-go credits from the tool's maker
Sources
Picked by hand (Phase 0); see the case study entry 'Whose communities?'.
168 sources across 22 projects, read on 2026-10-25. Each project page lists its own. Poll version: core-v4.