---
title: "Deploy herdr"
description: "herdr: persistent terminal workspace server for AI coding agents."
category: "AI/ML"
url: https://railway.com/deploy/herdr
---

# Deploy herdr

herdr: persistent terminal workspace server for AI coding agents.

**[Deploy herdr on Railway](https://railway.com/template/herdr)**

Machine-readable deploy manifest (JSON, validated by TemplateCI): https://railway.com/deploy/herdr/manifest.json

- **Creator:** wotonews
- **Category:** AI/ML

## Template content

### herdr

- **Image:** wotonews/herdr:v0.9.3-1

## Documentation

# Deploy and Host herdr with Railway

herdr is the runtime your coding agents live on: a terminal workspace manager that keeps agent sessions running on a server after you detach. Panes are marked working, blocked, or idle, layouts and agent sessions survive restarts, and agents can drive herdr through its CLI and socket API.

## About Hosting herdr

This template runs the pinned upstream herdr binary in headless server mode with `HOME` on a persistent volume at `/data` — the API socket, saved layouts, session state, and logs all survive redeploys. There is no web UI and no HTTP port: you attach from your own machine with the herdr client through `railway ssh`, and your terminal becomes the window onto the server's panes. The image ships git, curl, ripgrep, jq, tmux, and an SSH client so coding agents (Claude Code, Codex, OpenCode, …) have a sane environment; install anything else from inside a pane.

## Why Deploy herdr on Railway?

- **Always on:** agents keep working after you close the laptop; heartbeats and long builds never die with your SSH session.
- **Persistent sessions:** the volume preserves layouts, session state, and agent history across redeploys.
- **Private by construction:** no public port — access rides Railway's authenticated SSH.
- **One platform:** host the agent runtime next to the repos' CI, APIs, and databases it uses.

## Common Use Cases

- A shared, always-on terminal server for a team of coding agents (Claude Code, Codex, OpenCode).
- Long-running agent tasks you check in on from anywhere, never babysitting a terminal.
- Pairing multiple agents across panes with blocked/idle status at a glance.
- A scratch Linux box for agents that outgrows your laptop.

## Dependencies for herdr Hosting

- Docker image: official [`herdr` `v0.9.3`](https://github.com/herdrdev/herdr/releases/tag/v0.9.3) release binary via `wotonews/herdr:v0.9.3-1`
- Debian 12 + git, curl, ripgrep, jq, tmux, openssh-client (agent-friendly tooling)
- Your own agent CLIs (installed on demand inside panes)

### Deployment Dependencies

- Source: https://github.com/herdrdev/herdr (Apache-2.0)
- Docs: https://herdr.dev/docs/
- Local client: `curl -fsSL https://herdr.dev/install.sh | sh`

## Connect

```bash
# one-time: install the herdr client locally (see herdr.dev), then:
railway ssh            # opens a shell in the service
herdr                  # attaches the TUI to the running server
```

Detach with the tmux-style prefix (`ctrl+b q`) and reattach any time — panes keep running.

## Environment Variables

None required. The image bakes `HOME=/data` and `TERM=xterm-256color`. There is no auth surface to configure — access is Railway SSH.

## Local Setup

```bash
docker build -t wotonews/herdr:v0.9.3-1 .
docker run -d --name herdr -v herdr-data:/data wotonews/herdr:v0.9.3-1
docker exec -it herdr herdr   # attach a client inside the container
```




## Similar templates

- [Chat Chat](https://railway.com/deploy/-WWW5r) — Chat Chat, your own unified chat and search to AI platform.
- [stella](https://railway.com/deploy/stella) — Self-host stella with web, API, Postgres, Redis, and object storage.
- [Hermes Agent | OpenClaw Alternative with Dashboard](https://railway.com/deploy/hermes-agent-or-openclaw-alternative-wit) — Self-Hosted Hermes AI Agent for Telegram, Discord & Slack

Open this page in a browser: https://railway.com/deploy/herdr