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

# Quickstart

> Start the live graph watcher

## 1. Run Supermodel in your repo

```bash theme={null}
cd /path/to/your/repo
supermodel
```

If no config exists, `supermodel` opens the setup flow automatically. It authenticates you, confirms the repo, offers to install the Claude Code hook, and enables `.graph.*` sidecar files.

After setup, or on any already configured repo, the same command starts the live graph watcher:

```text theme={null}
[supermodel] [step:1] Building code graph
[supermodel] [step:2] Starting listeners
[supermodel] [step:3] Ready — listening on UDP :7734
```

Leave it running while you code. It refreshes from hook notifications; if you are not using a hook, run `supermodel --fs-watch` to watch local file changes directly. It writes graph sidecars next to source files and cleans generated sidecars when you stop it with `Ctrl+C`.

## 2. Tell your agent the graph files exist

```bash theme={null}
supermodel skill >> CLAUDE.md
# or AGENTS.md, .cursorrules, etc.
```

Use `>>` to append to existing instructions instead of replacing them. If the Supermodel block is already present, skip this step or remove the old block before running it again.

The prompt explains the naming convention (`Foo.py` → `Foo.graph.py`) and tells your agent to read the graph file before the source file.

## 3. Use the live graph

| Goal                                 | Command                                   |
| ------------------------------------ | ----------------------------------------- |
| Find unreachable functions           | `supermodel dead-code`                    |
| See what a change impacts            | `supermodel blast-radius path/to/file.go` |
| Codebase health report               | `supermodel audit`                        |
| Find usages of a symbol              | `supermodel find handleRequest`           |
| Token-efficient context for one file | `supermodel focus path/to/file.go`        |
| Print the full graph                 | `supermodel graph`                        |

These commands reuse the cached graph automatically. Pass `--force` to bypass the cache.

## 4. One-shot analysis

Use `analyze` when you want to build or refresh the graph once and exit instead of keeping the watcher running:

```bash theme={null}
supermodel analyze
```

`analyze` uploads the repo, runs call graph + dependency + domain analysis, caches the result locally by content hash, and writes `.graph.*` sidecar files next to source files. Add `--no-shards` to skip writing sidecars.

## 5. Configure (optional)

Settings live at `~/.supermodel/config.yaml`. Environment variables override file values.

```yaml theme={null}
api_key: smsk_live_...                       # or SUPERMODEL_API_KEY
api_base: https://api.supermodeltools.com    # or SUPERMODEL_API_BASE
output: human                                # human | json
shards: true                                 # set false (or SUPERMODEL_SHARDS=false) to skip writing graph files
```

## Next

See the [`analyze` command reference](/cli/commands/analyze) for one-shot flags and examples.