# AI use cases for Openharness (autonomous-ai/openharness) Repo: [autonomous-ai/openharness](https://github.com/autonomous-ai/openharness) | Use cases: 5 | Page: https://dedupedai.com/use-cases/autonomous-ai-openharness/ Harness is a command center that runs and watches coding agents and creative tool agents across your own machines, with a viewer for each craft. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Run coding agents on several machines in one window: https://dedupedai.com/use-cases/autonomous-ai-openharness/run-agents-across-machines/ - Build a circuit and compare two versions of it: https://dedupedai.com/use-cases/autonomous-ai-openharness/design-circuit-and-compare-traces/ - Pin a robot moment and test a disturbance: https://dedupedai.com/use-cases/autonomous-ai-openharness/tune-robot-motion-in-mujoco/ - Wrap your team's toolchain as a harness: https://dedupedai.com/use-cases/autonomous-ai-openharness/author-a-custom-harness/ - Perform a Strudel track and keep the WAV: https://dedupedai.com/use-cases/autonomous-ai-openharness/perform-and-keep-a-strudel-track/ ## How to run coding agents on several machines in one window with Openharness Repo: [autonomous-ai/openharness](https://github.com/autonomous-ai/openharness) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Link a laptop and a GPU box so Claude Code and Codex run in one command center. **Who it is for:** A developer with a laptop and a home GPU box who wants agents on both without opening ports or setting up SSH keys. **The problem:** Each machine has its own agents and terminals, so checking progress means switching between computers. **The outcome:** Agents on both machines appear in one app window, and you can open, read, and answer any of them. ### Steps 1. Download the Harness app for macOS or Linux from the Harness site. 2. On the second machine, run `curl -fsSL https://harness.autonomous.ai/cli/install.sh | bash`. 3. Run `harness login`, then `harness remote-password set`, then `harness start` on that machine. 4. In the app, open Machines and choose Link Machine, then enter the password you set. 5. Start a Claude Code or Codex harness on each machine from the app and check that both appear. ### Prompt to paste into your agent ```text In the openharness repo at autonomous-ai/openharness, follow the Get started section to link the machine to the desktop app. Confirm that the daemon runs with `harness start` and that appears under Machines. Report the machine name and the list of harnesses running on it. ``` ### Caveat Harness is in active development and the README does not say which operating systems the CLI supports beyond the desktop app. Risk, remote script: The install step pipes a script from the internet directly into bash, so read it first or pin a trusted version. Risk, credentials: The remote password you set controls access to the machine, so store it in a password manager and do not share it. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/autonomous-ai-openharness/run-agents-across-machines/ ## How to build a circuit and compare two versions of it with Openharness Repo: [autonomous-ai/openharness](https://github.com/autonomous-ai/openharness) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Use a circuit agent in CircuitJS to change a resistor and overlay the results. **Who it is for:** An electronics hobbyist or student who wants to test how a small change in a filter affects its output. **The problem:** Comparing two circuit versions by hand means rebuilding the schematic and rerunning the simulation each time. **The outcome:** You get two overlaid traces showing the measured difference, and you keep both versions. ### Steps 1. Open the Store in the Harness app and install the CircuitJS harness. 2. Choose New Harness and select CircuitJS in your workspace. 3. Ask the agent to build a simple filter and describe the goal in plain words. 4. Ask the agent to change one resistor value and overlay the new trace on the old one. 5. Keep both versions and note the measured difference. ### Prompt to paste into your agent ```text Using the CircuitJS harness from the openharness repo, build a simple RC filter for . Change one resistor, overlay the two traces, and report the measured difference in cutoff frequency. Verify that both circuit versions are saved in the workspace. ``` ### Caveat The README shows CircuitJS in a demo clip only, so the exact steps inside the viewer may differ from this outline. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/autonomous-ai-openharness/design-circuit-and-compare-traces/ ## How to pin a robot moment and test a disturbance with Openharness Repo: [autonomous-ai/openharness](https://github.com/autonomous-ai/openharness) | Difficulty: Advanced | Time: 1 hour | Basis: From the README Use the MuJoCo harness to push a simulated robot and compare the two outcomes. **Who it is for:** A robotics student or engineer who wants to see how a simulated robot reacts to a sudden force. **The problem:** Testing how a robot recovers from a push usually means writing a new simulation script for each test. **The outcome:** You get two branching futures from the same pinned moment, showing the robot's response to a 100 N push. ### Steps 1. Open the Store in the Harness app and install the MuJoCo harness. 2. Choose New Harness and select MuJoCo. 3. Ask the agent to run the robot and pin a moment in its motion. 4. Ask the agent to apply a 100 N push at that moment and show both futures. 5. Keep the branch that matches your test goal. ### Prompt to paste into your agent ```text Using the MuJoCo harness in the openharness repo, run the robot model and pin the moment at . Apply a 100 N push and show the two resulting futures side by side. Verify that the pinned moment is saved and both outcomes are listed. ``` ### Caveat The README describes this workflow only through a demo clip, and it does not say which robot models MuJoCo supports in Harness. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/autonomous-ai-openharness/tune-robot-motion-in-mujoco/ ## How to wrap your team's toolchain as a harness with Openharness Repo: [autonomous-ai/openharness](https://github.com/autonomous-ai/openharness) | Difficulty: Advanced | Time: 1 hour | Basis: From the README Copy the hello-world example, check it, and install it as your own harness. **Who it is for:** A team lead or tool builder who wants a coding agent to drive an internal tool with its own instructions and viewer. **The problem:** Agents have no built-in instructions or viewer for a specialized toolchain, so each session starts from scratch. **The outcome:** A custom harness that appears in the New Harness list with your instructions and a working viewer. ### Steps 1. Clone the repo with `git clone https://github.com/autonomous-ai/openharness.git` and run `cd openharness`. 2. Run `harness dsh install "$PWD/store/viewers/web-viewer" --link`. 3. Run `cp -R store/examples/hello-world ../my-harness`, then edit the harness.json and instructions for your tool. 4. Run `harness dsh check ../my-harness`, then `harness dsh install ../my-harness --link`. 5. Press ⌘N in the app and choose your harness to test it. ### Prompt to paste into your agent ```text In the openharness repo, copy store/examples/hello-world to ../, update its harness.json and AGENTS.md for , then run `harness dsh check ../`. Report the check result and confirm the harness installs with `harness dsh install ../ --link`. ``` ### Caveat The README gives the authoring steps but not the full harness.json schema, so check the linked authoring guide before you edit the file. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/autonomous-ai-openharness/author-a-custom-harness/ ## How to perform a Strudel track and keep the WAV with Openharness Repo: [autonomous-ai/openharness](https://github.com/autonomous-ai/openharness) | Difficulty: Beginner | Time: 30 minutes | Basis: From the README Bring voices in and out of a live Strudel track and save the best take. **Who it is for:** A musician or hobby producer who writes beats as code and wants to record a take without a full DAW. **The problem:** Live-coding music is hard to record cleanly, and good takes are easy to lose. **The outcome:** A WAV file of the track with the good sections marked, plus the code that made it. ### Steps 1. Open the Store in the Harness app and install the Strudel harness. 2. Choose New Harness and select Strudel. 3. Ask the agent to build a short beat and bring voices in and out. 4. Mark the sections you like, then ask the agent to export the WAV. ### Prompt to paste into your agent ```text Using the Strudel harness in the openharness repo, write a four-bar beat at BPM, add and remove two voices, and export the result as a WAV. Verify that the WAV file and the Strudel source are both saved in the workspace. ``` ### Caveat The README shows a demo clip for Strudel but does not describe the export steps, so confirm the WAV option in the viewer. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/autonomous-ai-openharness/perform-and-keep-a-strudel-track/