4RGE CLI

Agents that speak shell.

Install once, sign in with your 4RGE org, then run interactively (4rge) or fire one-shots from scripts (4rge run). Models are managed; tools stay on your machine.

Install CLI Read docs

$curl https://4rge.ai/install -fsS | bash

4RGED CLI - managed terminal session

4RGED CLI is ready. Type a request and press Enter to stream plans, tool calls, file reads, and shell output in this session.

Route to the right model

Frontier models from your policy — switched with a flag or slash command, not a settings panel.

Use in your preferred editor

Run from the terminal panel in VS Code, JetBrains, or any shell you already live in.

Write scripts that scale

Nightly doc updates, security reviews, and CI gates — same CLI, promoted to GitHub Actions or Make.

Learn more

CLI overview

When to use 4rge vs Desktop, and how managed login fits.

Read documentation →

Install and login

Install script, 4rge login, status, and first interactive session.

Read documentation →

Scripts and automation

4rge run, pipes, sessions, and what is (and is not) ready for CI today.

Read documentation →

Terminal agent, managed account

Interactive in a repo

cd into the project and run 4rge. Chat with a managed model while the agent edits and runs commands in that checkout.

Terminal-first agents

Install once, sign in with your 4RGE org, then run interactively (4rge) or fire one-shots from scripts (4rge run). Models are managed; tools stay on your machine.

Route to the right model

Frontier models from your policy — switched with a flag or slash command, not a settings panel.

Chat

Open the TUI

4rge in a repo — interactive agent with local tools.

List, rename, and continue local sessions. Use --continue when the next prompt depends on the last thread.
Gateway-routed models after login — bash, files, LSP, MCP, and skills on your machine.

Use in your preferred editor

Run from the terminal panel in VS Code, JetBrains, or any shell you already live in.

Install

Install the binary

curl https://4rge.ai/install -fsS | bash — then 4rge login.

4rge run "…" for non-interactive prompts. Pipe stdin in, capture stdout — after you have signed in once.
cd into the project and run 4rge. Chat with a managed model while the agent edits and runs commands in that checkout.

Agents that speak shell.

Your repo. Your shell. Managed models.
Interactive TUI or 4rge run for scripts — local files, bash, LSP, and MCP, with org-governed models after login.

Terminal-first agents

Agents that speak shell.

Install once, sign in with your 4RGE org, then run interactively (4rge) or fire one-shots from scripts (4rge run). Models are managed; tools stay on your machine.

shone install, any directory
CIoutput built for pipes and gates
repocontext from where you stand

Write scripts that scale

Nightly doc updates, security reviews, and CI gates — same CLI, promoted to GitHub Actions or Make.

Confirm

Check org and models

4rge status and 4rge models show account, org, and what you can run.

Your repo. Your shell. Managed models.
Interactive TUI or 4rge run for scripts — local files, bash, LSP, and MCP, with org-governed models after login.

One-shot from scripts

4rge run "…" for non-interactive prompts. Pipe stdin in, capture stdout — after you have signed in once.

Install

Install the binary

curl https://4rge.ai/install -fsS | bash — then 4rge login.

sh — one install, any directory
CI — output built for pipes and gates
repo — context from where you stand

Install. Login. Run.

The real command path from the 4RGED CLI.

Managed models, local tools

Gateway-routed models after login — bash, files, LSP, MCP, and skills on your machine.

Check org and models

4rge status and 4rge models show account, org, and what you can run.

Sessions you can resume

List, rename, and continue local sessions. Use --continue when the next prompt depends on the last thread.

Chat

Open the TUI

4rge in a repo — interactive agent with local tools.

Install. Login. Run.
The real command path from the 4RGED CLI.

Fire a one-shot

4rge run "…" or pipe a file into 4rge run.

Managed models, local tools

Check org and models

4rge status and 4rge models show account, org, and what you can run.

Gateway-routed models after login — bash, files, LSP, MCP, and skills on your machine.
cd into the project and run 4rge. Chat with a managed model while the agent edits and runs commands in that checkout.

Route to the right model

Frontier models from your policy — switched with a flag or slash command, not a settings panel.

Chat

Open the TUI

4rge in a repo — interactive agent with local tools.

List, rename, and continue local sessions. Use --continue when the next prompt depends on the last thread.
Gateway-routed models after login — bash, files, LSP, MCP, and skills on your machine.

Use in your preferred editor

Run from the terminal panel in VS Code, JetBrains, or any shell you already live in.

Install

Install the binary

curl https://4rge.ai/install -fsS | bash — then 4rge login.

4rge run "…" for non-interactive prompts. Pipe stdin in, capture stdout — after you have signed in once.
cd into the project and run 4rge. Chat with a managed model while the agent edits and runs commands in that checkout.

Write scripts that scale

Nightly doc updates, security reviews, and CI gates — same CLI, promoted to GitHub Actions or Make.

Confirm

Check org and models

4rge status and 4rge models show account, org, and what you can run.

Your repo. Your shell. Managed models.
Interactive TUI or 4rge run for scripts — local files, bash, LSP, and MCP, with org-governed models after login.

One-shot from scripts

4rge run "…" for non-interactive prompts. Pipe stdin in, capture stdout — after you have signed in once.

Install

Install the binary

curl https://4rge.ai/install -fsS | bash — then 4rge login.

sh — one install, any directory
CI — output built for pipes and gates
repo — context from where you stand

Sessions you can resume

List, rename, and continue local sessions. Use --continue when the next prompt depends on the last thread.

Chat

Open the TUI

4rge in a repo — interactive agent with local tools.

Install. Login. Run.
The real command path from the 4RGED CLI.

Fire a one-shot

4rge run "…" or pipe a file into 4rge run.

Managed models, local tools

Check org and models

4rge status and 4rge models show account, org, and what you can run.

Gateway-routed models after login — bash, files, LSP, MCP, and skills on your machine.
cd into the project and run 4rge. Chat with a managed model while the agent edits and runs commands in that checkout.

Agents that speak shell.

Route to the right model

Use in your preferred editor

Write scripts that scale

Learn more

CLI overview

Install and login

Scripts and automation

Terminal agent, managed account

Route to the right model

Use in your preferred editor

Agents that speak shell.

Write scripts that scale

One-shot from scripts

Install. Login. Run.

Sessions you can resume

Fire a one-shot

Route to the right model

Use in your preferred editor

Write scripts that scale

One-shot from scripts

Sessions you can resume

Fire a one-shot

Approvals

Transactions

Card user

Responses

Pipelines