> ## Documentation Index
> Fetch the complete documentation index at: https://docs.reliantlabs.io/llms.txt
> Use this file to discover all available pages before exploring further.
# Built-in workflows
> The workflows that ship with Reliant, what each is for, and how to choose one
Reliant ships with a set of ready-made workflows. A workflow decides how a chat proceeds: whether one agent just answers you, or a pipeline implements, checks and reviews, or several agents compete. You pick one when you start a chat, and you can read or copy any of them as YAML (see [Custom workflows](/workflows/custom-workflows)).
| Workflow | Use it when | Guide |
| - | - | - |
| **Agent** | Day-to-day work. One agent with tools that loops until it has answered. The default | [Workflows](/workflows/overview) |
| **Get It Right** | A change in a complex existing codebase that must pass lint, tests and build, and survive a review | [Get It Right](/guides/get-it-right) |
| **Best of Three** | A task with more than one reasonable approach, where you want to see three attempts and keep the best | [Best of Three](/guides/best-of-three) |
| **Build an App** | Start a new Forge app from a conversation, and get a preview link at the end | [Forge quick start](/features/deploy-and-host) |
| **Move to Forge** | Port an existing codebase into a Forge project, component by component | [Forge quick start](/features/deploy-and-host) |
| **Landing Page** | A polished landing page: scope, content pass, style pass, then a live preview | below |
| **Pitch Deck** | An investor deck from your company website, competitive research and a short founder interview | below |
| **Create a Workflow** | Build your own workflow from a conversation | [Create a Workflow](/guides/create-a-workflow) |
| **Import from Claude Code / Cursor** | Bring your instructions, MCP servers and skills across from another tool | [Import](/guides/import) |
## Landing Page and Pitch Deck
Both are pipelines in the same style as Get It Right. **Landing Page** starts with a scoping conversation that writes a plan file, then runs one review loop focused on *content* (is it saying the right thing, and is it accurate) and a second focused on *style* (hierarchy, responsiveness, brand consistency). It ends by serving the page and handing you a preview URL, after which the chat stays open so you can keep iterating. A page meant to go live is built as a Forge static site, so hosting is offered as the next step. **Pitch Deck** reads your website, researches competitors and the market in parallel, interviews you only for what it could not discover, writes and reviews each slide in parallel, then builds the slides and checks them visually from screenshots. Both end as an ordinary agent chat on the finished result.
## What they have in common
Every built-in takes the same toolbar inputs: a **model**, a **mode** (`auto` approves tool use, `manual` asks first, and some offer `plan`), and often **ask**, which pauses for your review between phases. When no one is watching, as in an [automation](/automations/overview), checkpoints resolve on their own instead of waiting.
Choose a **preset** to change how the agents inside behave (a UX specialist for front-end work, say); see [Presets](/workflows/presets).
This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.