> ## 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.

# Before a large Scout run

> Check Scout, run a small pass, review evidence, and then export or sync larger results.

Use this checklist before a large export, a citywide research pass, or a sync to
Atlas.

## 1. Check The Command

```bash theme={null}
scout --help
scout db path
```

## 2. Prepare Scout

```bash theme={null}
scout setup
scout config model
scout search connect
scout login --atlas-url https://atlas.rebuildingus.org
```

For scripts, pass credentials through the shell:

```bash theme={null}
export SEARCH_API_KEY=...
export ATLAS_API_KEY=...
```

## 3. Check Readiness

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

Do not start a large run until Scout is ready for the work you plan to do.

## 4. Run A Small Pass First

Start with one known public page:

```bash theme={null}
scout run \
  --prompt "Find source-backed civic actors and keep the source context." \
  https://example.org/public-page
```

Then test the place and issue set:

```bash theme={null}
scout run \
  --location "Las Vegas, NV" \
  --issues housing_affordability,public_transit \
  --target-count 25
```

## 5. Check Browser Rendering When Needed

If the sources depend on JavaScript, test a bounded crawl:

```bash theme={null}
scout articles crawl \
  --seed https://example.org/news \
  --target-count 5 \
  --max-pages 20 \
  --max-depth 1 \
  --browser-renders 3 \
  --json
```

Scout should use normal HTML first and keep browser renders within the budget.

## 6. Review The Results

```bash theme={null}
scout runs list
scout entries stats
scout entries list --type person --random --unique-names --limit 25 --format json
```

Check source title, URL, date, type, and context before export or sync. Do not
treat a name as current just because Scout found it.

## 7. Export Or Sync

Export reviewed people:

```bash theme={null}
scout export entries \
  --run-id RUN_ID \
  --type person \
  --unique-names \
  --format jsonl \
  --output people.jsonl
```

Sync a reviewed run:

```bash theme={null}
scout sync RUN_ID \
  --atlas-url https://atlas.rebuildingus.org \
  --target public
```

Keep uncertain work private to a workspace:

```bash theme={null}
scout sync RUN_ID \
  --atlas-url https://atlas.rebuildingus.org \
  --target workspace \
  --workspace WORKSPACE_ID
```

## Ready To Scale

A large run is ready when:

* Scout is ready for direct runs, search discovery, and sync.
* A small direct URL run and a small place-and-issue run both produce useful
  results.
* Browser rendering stays bounded when it is used.
* Sampled entries include source context that supports the claim.
* Exported files preserve source fields.
* Sync returns a receipt and does not directly publish public records.