I Built a Mobile Terminal Around My herdr + Codex Workflow

I am building Termish, an open-source mobile remote-work tool for SSH/Mosh terminals, file management, remote screens, and AI-assisted development.

The unusual part is that I built much of Termish through Termish itself.

My daily workflow is:

Termish on my phone → herdr on my host → Codex in the project

The development environment and AI tools run on my own machine. My phone is where I enter commands, describe tasks, upload screenshots, inspect changes, and check the resulting UI.

Termish is both the product I am building and a tool I use to build it.

Put remote work in your pocket.

Source code: github.com/ttermish/termish

Why do AI-assisted coding on a phone?

A phone is not a replacement for a desktop machine.

Its screen is smaller. Reading a large codebase is harder. Long debugging sessions are more comfortable on a larger display. Those limitations are real.

But AI-assisted development changes part of the workflow. Many tasks become:

1. Describe a requirement.
2. Provide context.
3. Let an agent make changes.
4. Review the output.
5. Run a command or test.
6. Decide what to change next.

Some of that loop works well on a phone.

For example, when I think of a feature away from my desk, I can open the project and ask an agent to start implementing it. When I find a UI problem, I can upload a screenshot to the host and ask the agent to inspect it together with the code. After the change is done, I can review the diff, run tests, and check the actual UI through a remote screen.

The goal is not to write an entire application on a phone. The goal is to make a phone a useful place to start work, inspect work, and keep a task moving.

Why another SSH and AI coding tool?

This is a fair question.

Tools such as Termius and Blink already provide strong SSH, Mosh, and file-management experiences. If you already have a setup that works well, you can run Codex, Claude Code, or other agents through a terminal today.

Happy and similar products also provide mobile interfaces for coding agents.

My own workflow moves constantly between:

• AI agents
• shell commands
• logs
• Git diffs
• project files
• screenshots
• the actual running application

I wanted a terminal-first experience that connects those steps on a phone.

Termish focuses on this sequence:

1. Connect to a host.
2. Prepare the remote tools.
3. Enter a herdr terminal workspace.
4. Work with Codex or another agent.
5. Upload a screenshot or inspect project files.
6. Review Git changes.
7. Open a remote screen and check the actual application.
8. Return to the terminal and continue.

The terminal remains the primary workspace. The AI workspace is there for people who prefer a native chat interface.

A terminal-first mobile workspace

Termish includes an SSH/Mosh terminal designed for touch devices.

It provides common terminal keys such as Ctrl, Esc, Tab, and arrow keys. It handles composition input for Chinese keyboards and can send touch input to compatible terminal applications. With a configured speech-recognition service, I can also dictate a task.

Small details matter on a phone.

If I upload a screenshot through the terminal, Termish can insert the remote file path into the terminal after the upload completes. I do not need to manually copy and paste it into an agent prompt.

If I open file management from a terminal session, Termish can jump directly to the current working directory.

None of these features is dramatic on its own. Together, they remove small interruptions that become much more noticeable on a phone.

Remote screen for checking the real result

Terminal output is not always enough.

After asking an agent to adjust an interface, I often need to see the running application. Termish includes a remote screen that can be opened full-screen or displayed as a smaller window alongside the terminal.

This lets me:

• inspect a running web or desktop application
• operate the remote machine when needed
• verify visual changes after an agent task
• return to the terminal and continue the task

A full remote desktop is useful when I need to work directly with a graphical interface. Most of the time, though, I want to stay in the terminal and open the remote screen only when I need to check the result.

Setup should not require sitting at a computer first

One experience I care about is avoiding a long manual setup process before a phone can become useful.

After you add a host that can be reached through SSH, Termish checks the remote environment when you use a related feature and can guide installation.

For example:

• If herdr is missing, Termish shows an installation card, runs the installation, displays logs, and lets you enter the workspace when it is ready.
• If Mosh is missing, Termish can guide installation and then continue with Mosh or SSH.
• If a supported agent CLI is missing, Termish can offer to install it.
• If the remote-screen service is missing, Termish provides an installation entry point.

This does not remove every setup step. System permissions and AI accounts still need explicit authorization. But the intention is that connecting to a host, preparing tools, and starting a task can continue from the phone.

Termish → herdr → Codex

The workflow I use most is simple:

• Termish is the mobile terminal, file manager, input layer, host list, and remote-screen entry point.
• herdr is the terminal workspace running on my host.
• Codex works inside the project and carries out development tasks.

This combination works well for me because the phone is only the interface. My code, credentials, terminal sessions, and agents remain on my own host.

If the phone disconnects, I can keep tasks alive with server-side tools such as herdr or tmux.

A native AI workspace is available too

I use herdr and Codex most often, but not everyone wants to operate through a terminal.

Termish also includes a native AI workspace. After connecting to a host, you choose an agent and a project directory, then describe a task through a mobile chat interface. You can send attachments, inspect execution progress, review results, and continue previous sessions.

The current supported agents are:

• Codex
• Claude Code
• Gemini CLI
• OpenCode
• Pi

The tasks still run on your own remote machine.

If you prefer a terminal, you can stay in herdr and use your existing command-line workflow. If you prefer a chat interface, you can start from the AI workspace.

Availability and project boundaries

Termish is MIT-licensed and available at:

• Website: termish.dev
• Source code: github.com/ttermish/termish

Android APKs are available now.

The iOS version can currently be built from source with Xcode, but it is not publicly distributed through the App Store yet.

Termish does not provide hosted development machines, relay your sessions, or include AI model subscriptions. You bring your own host, SSH access, and AI account or API.

What I would like feedback on

I would especially like to hear from people who access development machines from a phone.

Which part of the workflow is most painful today?

• Initial environment setup?
• Terminal input on a touch screen?
• Switching sessions?
• Moving screenshots and files?
• Reviewing agent output?
• Checking the running UI?
• Something else entirely?

I am interested in the workflows people already use: SSH + tmux, herdr, remote desktop, or mobile clients for coding agents.

If Termish fits your workflow, I would be glad to hear what is missing.
I am building Termish, an open-source mobile remote-work tool for SSH/Mosh terminals, file management, remote screens, and AI-assisted development.

The unusual part is that I built much of Termish through Termish itself.

My daily workflow is:

Termish on my phone → herdr on my host → Codex in the project

The development environment and AI tools run on my own machine. My phone is where I enter commands, describe tasks, upload screenshots, inspect changes, and check the resulting UI.

Termish is both the product I am building and a tool I use to build it.

Put remote work in your pocket.

Source code: github.com/ttermish/termish

Why do AI-assisted coding on a phone?

A phone is not a replacement for a desktop machine.

Its screen is smaller. Reading a large codebase is harder. Long debugging sessions are more comfortable on a larger display. Those limitations are real.

But AI-assisted development changes part of the workflow. Many tasks become:

1. Describe a requirement.
2. Provide context.
3. Let an agent make changes.
4. Review the output.
5. Run a command or test.
6. Decide what to change next.

Some of that loop works well on a phone.

For example, when I think of a feature away from my desk, I can open the project and ask an agent to start implementing it. When I find a UI problem, I can upload a screenshot to the host and ask the agent to inspect it together with the code. After the change is done, I can review the diff, run tests, and check the actual UI through a remote screen.

The goal is not to write an entire application on a phone. The goal is to make a phone a useful place to start work, inspect work, and keep a task moving.

Why another SSH and AI coding tool?

This is a fair question.

Tools such as Termius and Blink already provide strong SSH, Mosh, and file-management experiences. If you already have a setup that works well, you can run Codex, Claude Code, or other agents through a terminal today.

Happy and similar products also provide mobile interfaces for coding agents.

My own workflow moves constantly between:

• AI agents
• shell commands
• logs
• Git diffs
• project files
• screenshots
• the actual running application

I wanted a terminal-first experience that connects those steps on a phone.

Termish focuses on this sequence:

1. Connect to a host.
2. Prepare the remote tools.
3. Enter a herdr terminal workspace.
4. Work with Codex or another agent.
5. Upload a screenshot or inspect project files.
6. Review Git changes.
7. Open a remote screen and check the actual application.
8. Return to the terminal and continue.

The terminal remains the primary workspace. The AI workspace is there for people who prefer a native chat interface.

A terminal-first mobile workspace

Termish includes an SSH/Mosh terminal designed for touch devices.

It provides common terminal keys such as Ctrl, Esc, Tab, and arrow keys. It handles composition input for Chinese keyboards and can send touch input to compatible terminal applications. With a configured speech-recognition service, I can also dictate a task.

Small details matter on a phone.

If I upload a screenshot through the terminal, Termish can insert the remote file path into the terminal after the upload completes. I do not need to manually copy and paste it into an agent prompt.

If I open file management from a terminal session, Termish can jump directly to the current working directory.

None of these features is dramatic on its own. Together, they remove small interruptions that become much more noticeable on a phone.

Remote screen for checking the real result

Terminal output is not always enough.

After asking an agent to adjust an interface, I often need to see the running application. Termish includes a remote screen that can be opened full-screen or displayed as a smaller window alongside the terminal.

This lets me:

• inspect a running web or desktop application
• operate the remote machine when needed
• verify visual changes after an agent task
• return to the terminal and continue the task

A full remote desktop is useful when I need to work directly with a graphical interface. Most of the time, though, I want to stay in the terminal and open the remote screen only when I need to check the result.

Setup should not require sitting at a computer first

One experience I care about is avoiding a long manual setup process before a phone can become useful.

After you add a host that can be reached through SSH, Termish checks the remote environment when you use a related feature and can guide installation.

For example:

• If herdr is missing, Termish shows an installation card, runs the installation, displays logs, and lets you enter the workspace when it is ready.
• If Mosh is missing, Termish can guide installation and then continue with Mosh or SSH.
• If a supported agent CLI is missing, Termish can offer to install it.
• If the remote-screen service is missing, Termish provides an installation entry point.

This does not remove every setup step. System permissions and AI accounts still need explicit authorization. But the intention is that connecting to a host, preparing tools, and starting a task can continue from the phone.

Termish → herdr → Codex

The workflow I use most is simple:

• Termish is the mobile terminal, file manager, input layer, host list, and remote-screen entry point.
• herdr is the terminal workspace running on my host.
• Codex works inside the project and carries out development tasks.

This combination works well for me because the phone is only the interface. My code, credentials, terminal sessions, and agents remain on my own host.

If the phone disconnects, I can keep tasks alive with server-side tools such as herdr or tmux.

A native AI workspace is available too

I use herdr and Codex most often, but not everyone wants to operate through a terminal.

Termish also includes a native AI workspace. After connecting to a host, you choose an agent and a project directory, then describe a task through a mobile chat interface. You can send attachments, inspect execution progress, review results, and continue previous sessions.

The current supported agents are:

• Codex
• Claude Code
• Gemini CLI
• OpenCode
• Pi

The tasks still run on your own remote machine.

If you prefer a terminal, you can stay in herdr and use your existing command-line workflow. If you prefer a chat interface, you can start from the AI workspace.

Availability and project boundaries

Termish is MIT-licensed and available at:

• Website: termish.dev
• Source code: github.com/ttermish/termish

Android APKs are available now.

The iOS version can currently be built from source with Xcode, but it is not publicly distributed through the App Store yet.

Termish does not provide hosted development machines, relay your sessions, or include AI model subscriptions. You bring your own host, SSH access, and AI account or API.

What I would like feedback on

I would especially like to hear from people who access development machines from a phone.

Which part of the workflow is most painful today?

• Initial environment setup?
• Terminal input on a touch screen?
• Switching sessions?
• Moving screenshots and files?
• Reviewing agent output?
• Checking the running UI?
• Something else entirely?

I am interested in the workflows people already use: SSH + tmux, herdr, remote desktop, or mobile clients for coding agents.

If Termish fits your workflow, I would be glad to hear what is missing.