> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.islo.dev/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.islo.dev/_mcp/server.

# Quickstart with a coding agent

> Use the Islo CLI and factory-lines skill to have your coding agent create a Machine, Agents, and a Line, then inspect the first run.

Use this alternative to the [dashboard quickstart](/getting-started/quick-start) when you want your coding agent to prepare your [factory](/getting-started/glossary#factory). Your agent creates the [Machine](/getting-started/glossary#machine), [Agents](/getting-started/glossary#agent), and [Line](/getting-started/glossary#line) while you review the plan.

## Before you begin

Have a terminal, a local coding agent that supports skills, and a GitHub account with access to the repository you want to work on. You need an Islo account, available credits, and Machines enabled for your account. An organization owner may need to approve the GitHub app before a run can start.

#### Install the CLI and sign in

```bash
curl -fsSL https://islo.dev/install.sh | bash
islo login
```

#### Connect GitHub and install the skills

```bash
islo login --tool github
npx skills add islo-labs/skills
```

For ticket workflows, also connect [Linear](/integrations/linear) with `islo login --tool linear`. See [Authentication](/cli/authentication) for connection scope.

#### Ask your agent to prepare a Line

Replace the repository and PR number in this copyable prompt:

```text
Use the Islo factory-lines skill to create a manual Line that reviews a pull request in acme/api. Select my factory, prepare and verify a Machine, create the review Agent, validate the Line, and show me the deployment plan. After I review it, deploy the Line, start a run for pull request 123, and return the run ID.
```

The agent checks the installed `islo schema factory` and `islo schema agent` definitions while producing the files. Review its Agent, Machine, trigger, and [stage](/getting-started/glossary#stage) choices before deployment. The [factory model](/concepts/factory-model) explains factory selection; [Machines](/concepts/machines) explains build verification.

#### Inspect the run

Use the [run](/getting-started/glossary#run) ID your agent returns:

```bash
islo factory line-run status <run-id>
islo factory line-run events <run-id>
```

Status shows the current stage. Events show what happened at each transition. You can disconnect while agents work and return later. Open **Runs** in the dashboard and select the run for **Timeline** and **Chat**.

## What you'll see

Your agent returns a deployment plan and, after you approve deployment, a run ID. If the Machine is not ready, follow the agent's build verification instructions. For missing integrations, credits, or GitHub approval, see the [quickstart blockers](/getting-started/quick-start#if-something-blocks-you).

## Pick a goal to go deeper

#### [Fix a bug](/use-cases/bug-fix)

Turn a confirmed bug into a pull request.

#### [Deliver a feature](/use-cases/feature-delivery)

Ship a planned ticket with review and verification.

#### [Build your own Line](/use-cases/custom-lines)

Choose your Agents, stages, and triggers.

## Related resources

Read [How Islo works](/getting-started/how-islo-works), [Agents](/concepts/agents), [Machines](/concepts/machines), and [Triggers](/concepts/triggers) to tune the workflow. [Operate a run](/operate/runs) covers intervention, and [Delegations](/operate/delegations) covers requests from Slack, GitHub, and Linear after your factory is ready.