Quick Start

Sign in, set up a machine and a model, and run your first workflow. Setup takes a few minutes; most of it is choices, and you can change any of them later in Settings. Open the web app or launch the desktop app and sign in. Reliant then guides you through setup. The progress bar shows the steps: Machine, Model, Payment, Project and GitHub.

Machine

Reliant’s servers run your agents, but the agents do their work (reading files, running commands, running tests) on a machine you choose. See Key Concepts for how that fits together.

A Reliant cloud machine runs in Reliant Cloud. There is nothing to install, it starts for you, and it’s reachable from any device, including your phone. Machines come in five sizes, Compute Small through Compute 2XL, at $15, $35, $79, $159 and $299 per month. Your plan’s price is shown beside each size.
Use your own computer runs the agents’ tools on the computer you’re sitting at. In the desktop app, Reliant connects it for you. From the web app, setup shows the instructions for connecting it.

Model

Choose where the AI comes from:

Reliant is the managed option. You add credit to a wallet and pay as you go; there is no API key to find or paste.
A subscription you already have: sign in with Claude Code, Codex (ChatGPT), GitHub Copilot or Antigravity.
Your own API key for Anthropic, OpenAI or OpenRouter.

AI Providers covers each option, including what it costs.

Payment

This step only appears when something you chose costs money, a paid cloud machine or Reliant wallet credit. If you picked your own computer and your own provider, you skip it. If you have a coupon, you can redeem it here.

Project and GitHub

For a cloud machine, you choose where to start. Start something new has Reliant set up a project from scratch and get to work on your first idea. Connect GitHub lets you pick one of your repositories, and Reliant copies it onto your machine to work on it. You can add more projects later from Projects, so this is only where you start. If you chose your own computer, setup finishes after the model step. Add a project from Projects whenever you’re ready.

2. Start chatting

Once setup finishes you land in the app. The left sidebar starts with New chat, then Inbox (what needs you: an approval or a question from an agent, a run waiting for its machine, a failing automation, a finished run, a machine running low on disk), Workflows (the built-in and custom workflows, their runs and automations), Projects and Search, with your chats listed underneath. A project built with Forge also gets Apps and Observability directly under New chat. The panel on the right of a chat has Files, Changes, Processes, Tasks and Browser tabs. Click New chat, type a message and press Enter:

“Explain the structure of this project”
“What does the main file do?”
“Add input validation to the signup form”

This is the default agent workflow running. Your message goes to the model, the model calls tools on your machine (read files, search code, make edits, run commands), the results go back to the model, and it loops until the task is done. If your machine is a cloud machine that was asleep, it wakes automatically and the first message may take a little longer.

3. Run a built-in workflow

This is where Reliant gets different. Instead of manually driving each step, pick a workflow and let agents execute a defined process. Click the workflow selector in the chat input (it defaults to “Agent”) and try one of these:

Workflow	When to use it	What happens
Get It Right	A complex change in a large codebase	Implement → lint/test/build → review, undoing bad attempts, until the reviewer passes it
Best of Three	You want to compare approaches	Three agents implement in separate worktrees → a reviewer picks or merges the best
Build an App	You’re starting something new that will be deployed: an API, workers, a web or mobile app, a static site	Scope with you → scaffold with Forge → build → verify the running app
Create a Workflow	You have a process to automate, including one that should run on a schedule or an event	Scope with you → build any integrations → write and test the workflow

Select a workflow, type your request, and watch multiple agents collaborate on the task. The full list is on the home page.

4. Build your own workflow

Create a YAML file in your project to define a custom process:

.reliant/workflows/review-and-refactor.yaml

name: review-and-refactor
description: Reviews code, then refactors based on findings
presets:
  tag: agent

entry: [review]

inputs:
  model:
    type: model
    default: ""

nodes:
  - id: review
    type: workflow
    ref: builtin://agent
    thread:
      mode: inherit
    args:
      model: "{{inputs.model}}"
      mode: auto
      system_prompt: |
        Critically review the code the user points you to. Focus on:
        - Unnecessary complexity
        - Duplicated logic
        - Unclear naming or structure
        Be specific. Reference exact code.

  - id: refactor
    type: workflow
    ref: builtin://agent
    thread:
      mode: inherit
      inject:
        role: user
        content: |
          Based on your review above, refactor the code to address the
          issues you found. Make minimal, focused changes.
    args:
      model: "{{inputs.model}}"
      mode: auto

  - id: verify
    type: run
    command: npm test

edges:
  - from: review
    default: refactor
  - from: refactor
    default: verify

This workflow: criticizes the code → refactors based on its own critique → runs tests to verify nothing broke. Three steps, zero manual intervention. See Creating Custom Workflows for the full guide.

Execution modes

Control how much autonomy the agent has with Auto (full autonomy), Manual (approve each action), or Plan (read-only exploration). You can switch modes mid-conversation — a common pattern is to start in Plan mode to analyze, then switch to Auto to execute. See Execution Modes for details.

Optional: memory files

Memory files provide persistent context to the AI across conversations. They live on the machine your agents run on, so a cloud machine has its own, separate from your laptop’s. Global: ~/.reliant/reliant.md

# My Guidelines
- Always write tests
- Use descriptive variable names

Project-specific: reliant.md in your project root

# Project Context
- TypeScript/React project
- Uses Prisma for database

Learn more: Memories & Context

Next steps

Key Concepts — Understand the essential terms
Workflows Overview — How workflows power Reliant
Workflow Control — Pause, resume, and live parameter updates
Creating Custom Workflows — Build your own
Workflow Examples — Copy-and-adapt patterns and complete workflows

Get started

Use Reliant

Built-in workflows

Machines

Command line

Automations & integrations

Build workflows

Customize

Build & host with Forge

Account & billing

Reference

Help

Machine

Model

Payment

Project and GitHub

2. Start chatting

3. Run a built-in workflow

4. Build your own workflow

Execution modes

Optional: memory files

Next steps

​1. Sign in and walk through setup

​Machine

​Model

​Payment

​Project and GitHub

​2. Start chatting

​3. Run a built-in workflow

​4. Build your own workflow

​Execution modes

​Optional: memory files

​Next steps

1. Sign in and walk through setup

Machine

Model

Payment

Project and GitHub

2. Start chatting

3. Run a built-in workflow

4. Build your own workflow

Execution modes

Optional: memory files

Next steps