Our take
Newless than six months old and changing fast- Activity
- 82
- Community
- 54
- Issues
- 76
- Pull requests
- 84
+2,325 stars in the last 30 days
- Merges more pull requests than 96% of the projects we track
- Releases more often than 81% of the projects we track
- Closes issues faster than 81% of the projects we track
OMH is an ambitious, actively changing Hermes extension for developers who want parallel agent workflows and explicit verification, but its recent launch means its practices and integrations are still evolving.
Good fit if
- You already use Hermes Agent and want a separate layer for routing, delegation, and task graphs.
- Your coding work benefits from parallel lanes, specialist skills, and a distinct verification step.
- You can work with Python 3.11 or newer and are comfortable evaluating a young project under the MIT license.
Look elsewhere if
- You need an agent orchestration layer that works independently of Hermes Agent.
- You require a mature, long-established integration rather than a project that is less than six months old and changing quickly.
- Your workflow does not need parallel coding agents, model routing, or evidence-oriented completion reporting.
All health signals
| Last commit | 2026-11-11 (today) |
|---|---|
| Commits, last 90 days | 100+ |
| Releases, last 12 months | 33 (latest omher-v0.2.3, 2026-11-10) |
| Contributors | 40 (top contributor: 72% of commits) |
| Issues closed, last 90 days | 4+ (typically closed in 0 days) |
| Pull requests merged, last 90 days | 95 (typically merged in 0 days) |
| Project age | 5 months |
Checked on 2026-11-11 with the GitHub API.
Project history
Daily GitHub snapshots since Oct 7, 2026. History covers stars and forks; health signals are not tracked historically.
Stars
5,751Forks
469Change since Nov 17, 2026: +25 stars路0 forks
Overview
oh-my-hermes (OMH) is an independent operating layer built on Hermes Agent by Nous Research. It keeps Hermes as the agent you talk to, while routing work to AI coworkers with their own skills and model settings. Its central dag-ulw workflow turns a goal into a graph of small tasks that can run in parallel, then uses a separate node to verify the goal.
The project targets developers using Hermes for substantial coding work who want structured delegation, model routing, and clearer evidence about what actually ran. It also offers a desktop app for macOS and Windows, alongside a native terminal interface. OMH says it does not patch or replace Hermes.
Key Features
- Builds task graphs with
dag-ulw, running ready tasks in parallel and assigning each task a model category and reasoning effort. - Routes work through editable model chains, with fallback when a provider rejects a model.
- Provides more than a hundred specialist skills that its router can load into a run.
- Separates reported completion from verified results, using statuses such as
Code 路 reported doneandTest 路 verified. - Offers a native terminal interface for viewing runs, steering or interrupting agents, and inspecting task models and token usage.
- Supports optional language-server tools for code navigation and semantic renames.
- Includes a review-gated memory system that prepares task-relevant recall without reading or modifying Hermes' own memory.
Use Cases
- A developer with a large feature request can use
dag-ulwto split it into parallel tasks and request a separate goal check. - A team using several model families can assign different model chains and reasoning effort to different work categories.
- A Hermes user can delegate specialist coding tasks without manually choosing and invoking every skill.
- Developers who want completion claims tied to tests, reviews, or CI can use OMH's stage-based evidence reporting.
What you need
Detected in the repository
- Python >=3.11 (from pyproject.toml)
- A test suite and automated checks on GitHub Actions
License in plain words
MITpermissive
- Commercial use: yes
- Modify and redistribute: yes
- You must keep: the copyright and license notice
- Share your changes: no
A summary, not legal advice: the LICENSE file is what applies.
Getting Started
On macOS or Linux, install and run setup:
curl -fsSL https://omher.app/install | sh
omh setup
See the README for Windows, package-manager options, and setup details.
Alternatives
- codeg: Codeg coordinates agents through the Agent Client Protocol in a shared workspace, rather than assigning Hermes-specific workers to DAG-based coding lanes.
- cccc: CCCC focuses on persistent agent teams, durable messaging, and cross-machine coordination, rather than DAG-planned parallel coding work.
- agent-orchestrator: Agent-orchestrator supervises Git-isolated tasks and tracks PRs, CI, and reviews in a Kanban, rather than routing work through Hermes-specific lanes.
- Vicoa: Vicoa connects agent CLIs across desktop, mobile, and remote machines, rather than centering on Hermes Agent and its DAG workflow.
| Project | Language | License | Stars | Status |
|---|---|---|---|---|
| oh-my-hermes | Python | MIT | 5.4k | New |
| codeg | Rust | Apache-2.0 | 4k | Active |
| cccc | Rust | Apache-2.0 | 1.3k | Active |
| agent-orchestrator | Go | Apache-2.0 | 13.3k | Active |
| Vicoa | Python | AGPL-3.0 | 516 | New |
Considerations
- OMH depends on Hermes Agent rather than replacing it, so adopting OMH also means adopting that underlying agent.
- The CLI requires Python 3.11 or newer. The README also describes desktop builds for macOS and Windows.
- The project is new and changing fast. Recent activity and releases are frequent, but this short history provides limited evidence of long-term stability.
- The README says the product A/B comparison has no measured run published yet, so it does not establish measured gains over Hermes alone.
- OMH is MIT-licensed; review the license when distributing or incorporating it.
- Project: GitHub repository 路 Documentation and website.
Found this useful?
Share it with someone who would like oh-my-hermes.