1. Made by
“claude plugin marketplace add autonomous-ai/openharness --sparse .claude-plugin plugins”
The repository autonomous-ai/openharness belongs to Autonomous, the company behind harness.autonomous.ai, and the vendor's own agents page installs the plugin from that repository.
harness.autonomous.ai/agents2. Keys
“claude mcp add harness -- npx -y @autonomous-ai/harness-mcp”github.com/autonomous-ai/openharness/blob/f5644bec684e0b98c433acb0253e64674c7ee06e/cli/packages/harness-mcp/README.md
4. Commercial use
“to use, copy, modify, merge, publish, distribute, sublicense, and/or sell”
MIT license; the tool can be used for paid work.
github.com/autonomous-ai/openharness/blob/f5644bec684e0b98c433acb0253e64674c7ee06e/LICENSEChecked against its repository at one commit, on Nov 10, 2026. Tap a verdict to see its source.How we checkedReport an error
What Harness does
Harness is a free app that runs coding assistants such as Claude Code and Codex, each in its own terminal window you can watch. With this server your assistant can start another assistant on a task, wait for it and read its answer. You see every one in the Harness app and can answer or stop it.
What people use it for
A second opinion from another model
While Claude Code builds your feature, it hands a review of the changes to Codex and reads the findings when they are ready.
Several tasks at the same time
Your assistant splits the work, for example fixing tests in two parts of the project, and starts one helper for each part. It waits for all of them, or for the first one to finish.
Long jobs that outlast the chat
A slow job, like a long test run, keeps going after your assistant's turn ends, and its answer waits to be read later.
Use your other computer
Your assistant on the laptop starts a job on your home server or a computer with a strong graphics card, once you have signed in to Harness on both.
Find what was already decided
Before starting new work, your assistant searches earlier conversations on this computer for what was said about the same problem.
Harness popularity
Tools it gives your agent (12)
What it sends and its license
the task text and the files each started agent reads
To that agent's own provider (for example Anthropic for Claude Code, OpenAI for Codex); every time a harness is started or given work. Opt out: do not start harnesses; the agent does the work itself.
terminal traffic between your machines, end-to-end encrypted
To directly between machines, or through Cloudflare TURN or the Harness relay when a direct path is blocked; only when you sign in and use harnesses on another machine. Opt out: keep harnesses on this computer; no account is needed then.
The app and this server are open source under the MIT license, so you can use them for paid work. Keep the license notice if you pass the code on. The agents you start bill your own Claude or OpenAI plans.
MIT license; the tool can be used for paid work.
“to use, copy, modify, merge, publish, distribute, sublicense, and/or sell” source
MIT license, no limit on who you do the work for.
“Permission is hereby granted, free of charge, to any person obtaining a copy” source
Passing the code on is allowed with the MIT copyright and permission notice kept; bundled upstream tools keep their own licenses.
“MIT license; upstream tools keep their own.” source
Is it alive
0.4.20 in the registry, updated Oct 19, 2026
npm @autonomous-ai/harness-mcp 0.2.0 (2026-10-19); the registry lists 0.4.20 with harness-mcp.mcpb in release v0.4.20_cli
Alternatives
Harness This page
Pick it if you want agent swarms with shared memory and hooks instead of watched sessions
io-github-plastic-labs-honcho
Pick it if you need memory that carries across your agent's sessions
Get started
Tell your AI assistant
“Use the Harness tools to start a Codex reviewer on this branch, wait for it, show me its findings and close it; call harness_whoami first and ask me before changing any Harness setting.”
Connect it
claude mcp add harness -- npx -y @autonomous-ai/harness-mcpnpx -y @autonomous-ai/harness-mcp{ "mcpServers": { "harness": { "command": "npx", "args": ["-y", "@autonomous-ai/harness-mcp"] } } }Mistakes to avoid
Each harness spends the person's tokens.
A folder Harness did not make makes Claude Code or Codex ask the person to trust it, and the harness waits.
Refusals are final and settings are the person's consent to give.
The wait is kept short so clients such as Codex do not end the call; the vendor page still says 120 seconds.
Two agents on one file overwrite each other's work.
Questions
Do I need an account?
Not for harnesses on this computer. You sign in only to let agents start harnesses, to use your other machines and the phone app.
What happens if Harness is not installed?
The server still starts but offers only one tool, harness_setup, which explains how to install Harness.
Can my assistant close or take over harnesses I started myself?
No. An agent may give work to, and close, only the harnesses it started.
Is the conversation lost when a harness is closed?
No. It is kept and you can resume it.