> ## Documentation Index
> Fetch the complete documentation index at: https://atlas.rebuildingus.org/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Set up Scout

> Connect Scout to Atlas, configure a model provider, and check readiness.

`scout setup` is the normal first-run path. It helps choose a local profile,
log in when needed, configure a local model provider, and install optional shell
integration.

```bash theme={null}
scout setup
```

## Atlas Login

Login connects this computer to Atlas for sync and worker mode:

```bash theme={null}
scout login
```

Scout opens an Atlas approval page in your browser. If you are on a server or
inside a terminal-only environment, print the approval URL instead:

```bash theme={null}
scout login --no-browser
```

Check the current login:

```bash theme={null}
scout auth status
scout whoami
```

Scout stores login tokens in your operating system's credential store. Atlas
does not store login tokens in Scout profile TOML files.

Local direct URL runs do not require login. Sync and worker mode do.

## Local Model Provider

Scout can use local providers such as LM Studio or Ollama. Configure the active
provider and model:

```bash theme={null}
scout config model
```

For an interactive repair flow:

```bash theme={null}
scout config model --interactive
```

For scripts, pass the provider and model explicitly:

```bash theme={null}
scout config model --provider lmstudio --model local-model-name
```

Use `scout config show` to inspect the active profile and `scout config path` to
print the profile file path.

## Profile Configuration

Scout profile files store non-secret settings for this computer. Use dedicated
commands for credentials and structured state:

```bash theme={null}
scout config show
scout config path
scout config get llm.provider
scout config set scraper.max_concurrent_fetches 6
```

`scout config set` accepts known scalar settings only. It will not save API
keys, tokens, or structured values.

Scheduled discovery targets use their own commands:

```bash theme={null}
scout config schedule set --enabled --cron "0 8 * * 1"
scout config schedule target add --location "Austin, TX" --issues housing_affordability,transportation
scout config schedule target list
```

## Search-Backed Discovery

Known URL runs and local article-corpus recovery do not need a search key.
Connect search when you want Scout to find new public sources for place and
issue discovery:

```bash theme={null}
scout search connect
scout search status
```

Scout stores saved search keys in your operating system's credential store.
`SEARCH_API_KEY` can be used as a script override for a single shell or
script.

Disconnect stored search credentials:

```bash theme={null}
scout search disconnect
```

## Readiness Check

Run doctor before a large run:

```bash theme={null}
scout doctor
```

Check the exact Atlas environment before sync:

```bash theme={null}
scout doctor --atlas-url https://atlas.rebuildingus.org
```

When a saved login is present, doctor contacts Atlas and verifies that the
login can create an upload token for the selected environment. It discards that
token and does not sync any run data.

Include worker checks only when this computer should process Atlas-assigned
jobs:

```bash theme={null}
scout doctor --worker
```

For scripts:

```bash theme={null}
scout doctor --json
```

For scripts, require the checks the run needs:

```bash theme={null}
scout doctor \
  --atlas-url https://atlas.rebuildingus.org \
  --require direct-url-runs \
  --require search-discovery \
  --require atlas-sync \
  --json
```

See [Readiness checks](/docs/scout/readiness) for what each check means.

## Profiles

Profiles keep laptop, workstation, and test settings separate:

```bash theme={null}
scout config create-profile studio
scout config use-profile studio
scout --profile studio run https://example.org
```