aq.dev / guides / coding-agent-harness-directory

The Coding Agent Harness Directory (November 2026)

This is a directory of coding agent harnesses and the tools that orchestrate them, checked against each vendor's own site in November 2026. It covers four shapes: CLIs built by model labs, open-source model-neutral harnesses, desktop apps that run several harnesses in parallel, and cloud platforms that run agents on hosted machines. It also lists what shut down or was replaced this year, the part every static awesome-list gets wrong: two well-known entries died in February 2026 alone. For the definition of the term, read what is a coding harness; for a decision framework, read how to choose one. This page is the map.

What changed on November 20, 2026: new models in Claude Code, Codex, and Antigravity; Cursor's weekly usage pool; Muse Code, Grok Bot, and Devin Bots added; Google Jules moved to the shutdown list; Conductor's reversal of its shutdown; HumanLayer joining Factory; Superset renamed Paradev; and current versions of Kimi Code and OpenCode.

How to read this directory

Four facets separate these tools better than feature lists:

  • Execution locus. Where the agent runs: your laptop, a VM you own, or the vendor's cloud. This decides who holds your code and credentials.
  • Model access. A subscription sign-in (Claude, ChatGPT), your own API keys, or local open-weight models.
  • Multiplayer access. Whether a second person can see or join a running session.
  • Persistence. Whether the session survives a closed laptop or a reboot.

CLIs from the model labs

Each lab now ships its own harness, tuned for its own models. All of these run locally in your terminal by default, so persistence is your problem (see keeping sessions alive past a closed laptop).

HarnessBuilt byModel accessNotes (as of November 2026)
Claude CodeAnthropicClaude Pro, Max, or Ultra plan, Team or Enterprise seat, or API keyTerminal, IDE extensions, desktop app, the web, and the Claude mobile app; macOS, Linux, Windows. Opus 5.5 is the default model on every plan; cloud sessions start from the browser, the mobile app, the desktop app, or claude --cloud; agent teams are experimental and off by default
Codex CLIOpenAIChatGPT plan sign-in or API keyOpen-source (Apache-2.0) terminal agent; one account also covers Codex in ChatGPT, the editor, and the cloud. GPT-6.2 Sol is the default model and GPT-6.2 Astra is for the hardest work (November 17, 2026). OpenAI says that starting in December, Plus, Pro, and Business allowances are shared across ChatGPT and Codex
Antigravity CLIGoogleGoogle account plansThe agy command; replaced Gemini CLI, which Google shut down for free, Pro, and Ultra users on June 18, 2026; multi-agent by default, with skills, hooks, and MCP support. Gemini 4 Neon arrived on November 18, 2026, with Orchestrator, where Neon plans and reviews and Gemini 4 Hydrogen implements; Remote Agents run sessions in Google's cloud for Google AI Pro and Ultra subscribers
Grok BuildxAIxAI account or API keyAnnounced May 25, 2026 as an early beta; open-source (Apache-2.0) TUI in Rust, with plan mode, headless CI use, and editor embedding over ACP
Kimi Code CLIMoonshot AIKimi models by default; other compatible providers configurableMIT-licensed TypeScript agent with built-in coder, explore, and plan subagents; succeeds the earlier kimi-cli. Version 2.7.2 (November 20, 2026) runs the K3.1 Pro models, and Kimi Code Desktop runs on macOS, Windows, and Linux
Muse CodeMetaMuse Code subscriptionMeta's coding agent for the terminal and CI, built for Muse Spark; out of beta since August 31, 2026; version 1.4.24 on Meta's stable update channel
AmpAmp (ampcode.com)Amp accountRuns on the web, in the terminal, in desktop apps for Mac, Windows, and Linux, and on iPhone, iPad, and Android; threads run in orbs, per-thread machines Amp hosts or you self-host, and workspace members can join a thread while it runs; bring your own Claude plan or keys with no Amp token fees; see AQ vs Amp

Open-source, model-neutral harnesses

These decouple the harness from the model vendor: bring an API key, a subscription, or a locally served open-weight model.

HarnessShapeNotes (as of November 2026)
OpenCodeTerminal, IDE, or desktop appConnects to any provider with your own keys, including local models; optional OpenCode Go subscription ($10 per month) for hosted open-source models. OpenCode 2 is an explicit upgrade from OpenCode 1 (opencode upgrade --major); the latest release is 2.3.2 (November 18, 2026)
PiMinimal terminal harnessShips four tools and a system prompt under 1,000 tokens; everything else arrives as extensions, skills, and packages shared over npm or git
AiderTerminal pair programmerAmong the oldest harnesses still in wide use; connects to most LLMs including local ones and git-commits every change
GooseLocal agent frameworkStarted at Block, now under the Agentic AI Foundation; model-configurable, extends through MCP
OpenHandsCLI, browser client, and agent serverMIT-licensed and model-agnostic; runs locally, self-hosted, or on its managed cloud

Desktop apps that run harnesses in parallel

These are orchestrators rather than harnesses: they run the stock CLIs side by side on your machine, usually one git worktree per task. This layer, the harness of harnesses, is where churn is fastest.

ToolPlatformsNotes (as of November 2026)
ConductormacOS, plus an iOS appRuns Claude Code, Codex, Cursor, and OpenCode in parallel, each workspace in its own worktree. Said on November 4, 2026 that it was joining Meta, then on November 13 that it is not shutting down; it keeps shipping releases
EmdashmacOS, Windows, LinuxFree and open source (YC W26); one worktree per task, works with 25+ CLI agents. The desktop app is now Emdash Classic and still gets fixes; new development moved to Emdash 2.0, cloud coding agents for teams
Superset (now Paradev)Mac, experimental Linux build, iOSRenamed Paradev; runs many coding agents in parallel on your machine, with a CLI, SDK, and MCP server
ProliferateDesktop, self-hostableOpen-source IDE for Claude Code, Codex, and OpenCode; runs agents in parallel, locally or in the cloud
XirpWeb dashboard over local sessionsSpotify's vendor-neutral agentic development environment, opened as a free public beta on August 10, 2026; manages Claude Code, Codex, and Gemini CLI sessions, one working tree per session
Vibe KanbanDesktop, open sourceKanban board over parallel agents with automatic worktrees; its company, bloop, shut down on April 10, 2026, and the project continues as community-maintained open source

Cloud platforms that run agents on hosted machines

Here execution moves off your machine: the vendor provisions a sandbox, runs the agent, and hands back a pull request.

PlatformVendorNotes (as of November 2026)
Claude Code on the webAnthropicClaude Code cloud sessions in Anthropic-managed sandboxes, started from the browser, the mobile app, the desktop app, or the CLI; on Pro, Max, Ultra, Team, and eligible Enterprise seats
Codex cloudOpenAIRuns many tasks in parallel, each in its own cloud sandbox preloaded with your repository
Cursor cloud agentsCursorAgents in isolated VMs or on machines you manage; hand a CLI conversation to the cloud by prefixing a message with an ampersand, resume on web or mobile. Every Cursor plan now draws on one usage pool shared across Cursor and Grok that refreshes weekly, and since November 5, 2026 apps built in Cursor can be published and shared with a link
Copilot coding agentGitHubAssign an issue to Copilot; it boots a GitHub Actions VM, writes the code, and opens a PR
DevinCognitionAutonomous engineer with a sandboxed VM per session; runs many tasks in parallel, with Devin Desktop, a CLI, and mobile apps. Devin Bots (Max and Teams plans) take an outcome to own, plan the work, and start Devin sessions until it is done
Grok BotxAIA persistent agent on a cloud computer, included on Premium, Super, and Ultra; its usage draws from the weekly pool shared across Grok, Grok Bot, and Cursor, and it stops when that pool runs out unless on-demand usage is on
Warp OzWarpLaunched February 2026; cloud agents in Docker sandboxes, driven by Claude Code, Codex, or Warp's own harness. Warp now calls cloud agents its Automation Platform, and the Oz web app, CLI, and APIs still serve existing workflows; Warp Factories (August 18, 2026) are infrastructure for building an internal software factory
SuperconductorSuperconductorCloud workspace running Claude Code, Codex, Pi, Amp, and others in parallel, with previews and one-click PRs from phone or desktop

Gone this year: the churn a directory has to track

This market retires tools faster than lists update. Status checked on each vendor's own site in November 2026:

  • Terragon shut down on February 9, 2026; its own docs now point users at Claude Code on the web.
  • Crystal was deprecated on February 26, 2026 and receives no further updates; its maker, Stravu, ships Nimbalyst as the successor.
  • Gemini CLI was shut down for free, Pro, and Ultra users on June 18, 2026, replaced by the Antigravity CLI.
  • kimi-cli was succeeded by Kimi Code CLI, which migrates its configuration and sessions.
  • Vibe Kanban lost its company: bloop shut down on April 10, 2026, and the project continues as community-maintained open source.
  • Google Jules went read-only after November 11, 2026 and winds down on December 31, 2026, according to its own changelog.
  • HumanLayer is joining Factory (announced November 4, 2026); the hosted platform stays up during a transition period, and HumanLayer is being open sourced.
  • Superset was renamed Paradev, and the open-source Emdash desktop app became Emdash Classic as development moved to Emdash 2.0.

One correction to the record: Conductor said on November 4, 2026 that it was joining Meta and would shut down its cloud, then said on November 13 that it is not shutting down. It stays in the desktop table above.

One adjacent entry worth knowing: qm, the MIT-licensed multiplayer agent harness Y Combinator open-sourced in July 2026, gives every employee a work agent over Slack and the web, driven by Pi, OpenCode, Codex, or Claude Code at its core. It is a company-wide work harness rather than a coding harness, and a sign of where the vocabulary is heading.

The facet the directory keeps surfacing: who sees the work

Scan the tables and a pattern shows: nearly everything local is single-player, and nearly everything multiplayer runs on the vendor's cloud. The default matters: a July 2026 LeadDev analysis of 25,264 agent-generated pull requests across 2,361 popular GitHub repositories found that in 79 percent of agentic PRs the same developer both reviewed and modified the agent's contribution, and only about one in eight workflows involved multiple humans. When a session is invisible, the only reviewable artifact is the diff, and the person closest to the diff is the person who prompted it. The missing quadrant, multiplayer sessions on infrastructure you own, is what multiplayer coding agents names.

Where AQ fits

AQ is the multiplayer coding harness where engineering teams run AI coding agents like Claude Code and Codex together: shared live terminals, a code editor, and app previews, in your own cloud. In this directory's facets, AQ sits in that missing quadrant. It is not another harness: agents run on the subscriptions and CLIs you already have (Claude Code, Codex, Cursor, Kimi Code, Grok, Antigravity, Devin, Muse Code, OpenCode, and more) on your team's machine, streamed live to the browser. Sessions survive a closed laptop and resume from any device, each workspace gets one isolated git worktree, and teammates open the same workspace and watch the same live session (typing into someone else's terminal happens only after its owner approves a control request in one click; visibility is owner-managed, team-visible or private and shared with specific people).

Model access stays as the tables describe: each engineer signs into the CLIs with their own Claude or OpenAI account, and AQ never marks up usage on your own subscriptions. Agents commit, push, and open PRs with per-user GitHub auth, every workspace can serve a live dev-server preview with links that are viewable without an account, and Linear issues become workspaces by labeling them ai-task, with a human starting the agent, never an auto-launch.

Pricing is two plans: Free runs agents on your own Mac with the AQ Mac app (private to you, free with no time limit), and AQ Pro is $50 per user per month in early access (standard $200, billed monthly), covering VMs you connect from your own cloud or a dedicated always-on AQ-managed VM in its own isolated network, with your rate locked for your first 12 months.

Plainly: pick any harness from the tables above, or several. AQ is the layer that makes whichever ones your team picks persistent, isolated, and visible to more than one person, without moving execution into someone else's cloud.

Frequently asked questions

What is a coding agent harness?

The software wrapped around a language model that turns it into a working coding agent: the system prompt, tool definitions, context management, execution loop, and failure handling. Claude Code, Codex CLI, OpenCode, and Pi are all harnesses; the model underneath supplies the reasoning. The glossary page covers the term in depth, including why the same model scores differently in different harnesses.

Which coding agent harnesses are open source?

As of November 2026: Codex CLI, OpenCode, Pi, Aider, Goose, OpenHands, Kimi Code CLI, and Grok Build among the harnesses themselves, plus Emdash Classic, Proliferate, and Vibe Kanban among the tools that orchestrate them. Claude Code, Amp, and the cloud platforms (Devin, Warp Oz, Superconductor) are proprietary. Open source here usually means the harness code; the frontier models it drives remain hosted APIs unless you serve open-weight models yourself.

Which coding agent tools shut down or were replaced in 2026?

Terragon shut down on February 9, 2026, and Crystal was deprecated on February 26, 2026 (its maker now ships Nimbalyst). Google shut down Gemini CLI for free, Pro, and Ultra users on June 18, 2026, replacing it with the Antigravity CLI, and Moonshot AI succeeded kimi-cli with Kimi Code CLI. Vibe Kanban's company shut down in April and the project continues as community-maintained open source. Google Jules went read-only after November 11, 2026 and winds down on December 31, HumanLayer is joining Factory, and Superset was renamed Paradev. Conductor said it was joining Meta, then on November 13 said it is not shutting down. Check the vendor's own site before adopting anything in this category.

Can I run multiple harnesses against the same repository?

Yes, and it is increasingly the norm rather than the exception. The safe pattern is one git worktree per session, so parallel agents never edit the same checkout: desktop orchestrators like Conductor and Emdash automate exactly that on one machine, and AQ does it on a team VM, where each workspace is an isolated worktree and any CLI can run inside it. Keep shared instruction files (CLAUDE.md, AGENTS.md) consistent across the harnesses you use.

Which coding harness should I choose?

Start from your model access: an Anthropic subscription points at Claude Code, a ChatGPT plan at Codex CLI, own keys or local models at OpenCode or Pi. Then run a small bake-off on your own repository rather than trusting benchmarks, because scores are earned by a model and harness pairing, not a model. The chooser guide walks the full decision, and treats the harness as a per-engineer pick you revisit as models change.