Oh My Hermes: One command to arm Hermes Agent into an all-round workstation - long-term memory, intelligent routing, and HUD panel all installed at onceOh My Hermes：一条命令把 Hermes Agent 武装成全能工作站——长期记忆、智能路由、HUD 面板一次装齐

Parallel workflow ulw-work: A task is automatically split into multiple non-conflicting lanes, each of which is bound to an independent model and started working on the git work tree at the same time, and finally merged into one merge.

Intelligent routing: Score requests according to task complexity. Handle simple problems to cheap models, and only upgrade to flagship models for difficult problems, saving money without sacrificing quality.

Long-term memory system: session records are automatically layered into the database, and the most relevant memory packages are recalled according to the budget for the next session. The more you use it, the better you will understand your project.

OMH HUD panel: The model, gear, round, token consumption and cost source of each lane are visible in real time, so you know how long you run.

The launch banner will list available tools, group skills, OMH experts and model pools on one screen, so you can see at a glance what capabilities you have.

When running ulw-work, the panel displays the plan and the status of each lane in real time; when three independent tasks are executed in parallel, all tool calls are recorded.

macOS/Linux one-line command: curl -fsSL https://raw.githubusercontent.com/rlaope/oh-my-hermes/main/install.sh | sh; Windows PowerShell: irm https://raw.githubusercontent.com/rlaope/oh-my-hermes/main/install.ps1 | iex.

You can also use the wheel package of this site for offline installation: pip install oh_my_hermes-3.0.0-py3-none-any.whl. After installation, run omh doctor to self-check the environment.

After installation, restart the Hermes Agent session and enter a skill list command such as /skills to confirm that the omh-* expert skills have been loaded.

The download button provides the v3.0.0 official wheel installation package (about 5.5 MB) from GitHub rlaope/oh-my-hermes Releases, which can be installed with one pip command.

Oh My Hermes requires the Hermes Agent environment to be used; parallel lane will call multiple models at the same time, so pay attention to the API quota consumption. This article is compiled from public information. The resources are for learning and exchange only. Please follow the source permission.