> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.nvidia.com/nemo/guardrails/llms.txt.
> For full documentation content, see https://docs.nvidia.com/nemo/guardrails/llms-full.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.nvidia.com/nemo/guardrails/_mcp/server.
# NeMo Guardrails Library CLI Reference
> Use the nemoguardrails command-line interface to chat, serve, evaluate, and convert configurations.
# NeMo Guardrails Library CLI Reference
You should now be able to invoke the `nemoguardrails` CLI.
```bash
> nemoguardrails --help
Usage: nemoguardrails [OPTIONS] COMMAND [ARGS]...
Options:
-v, --version Show version and exit.
--install-completion Install completion for the current shell.
--show-completion Show completion for the current shell, to copy it
or customize the installation.
--help Show this message and exit.
Commands:
chat Start an interactive chat session.
server Start a NeMo Guardrails server.
convert Convert Colang files and configs from older version to
the latest.
actions-server Start a NeMo Guardrails actions server.
find-providers List and select LLM providers interactively.
eval Evaluation a guardrail configuration.
```
You can also use the `--help` flag to learn more about each of the `nemoguardrails` commands.
***
## `chat`
```sh
nemoguardrails chat --config examples/configs/content_safety [--verbose] [--streaming]
```
### Options
| Option | Description |
| -------------------- | --------------------------------------------------------------------------------------------------------------------- |
| `--config` | Path to a directory containing configuration files. Can also point to a single configuration file. Default: `config`. |
| `--verbose` | Enable verbose mode with detailed logging information. |
| `--verbose-no-llm` | Enable verbose mode but exclude LLM prompts and responses. |
| `--verbose-simplify` | Simplify the verbose output. |
| `--debug-level` | Enable debug mode with rich flow execution info. Levels: `WARNING`, `INFO`, `DEBUG`. |
| `--streaming` | Enable streaming mode if the configuration supports it. |
| `--server-url` | Connect to a server instead of loading config locally. Requires `--config-id`. |
| `--config-id` | The config ID to use when connecting to a server. |
## `actions-server`
Start a separate server for running custom actions in an isolated environment.
```bash
nemoguardrails actions-server --port 8001
```
### Options
| Option | Description |
| -------- | ----------------------------------------------------------- |
| `--port` | The port that the server should listen on. Default: `8001`. |
***
## `server`
Start the FastAPI server for serving guardrails configurations over HTTP.
```bash
nemoguardrails server --config examples/configs --port 8000
```
### Options
| Option | Description |
| --------------------- | ------------------------------------------------------------------------- |
| `--port` | The port that the server should listen on. Default: `8000`. |
| `--config` | Path to a directory containing multiple configuration sub-folders. |
| `--default-config-id` | The default configuration to use when no config is specified in requests. |
| `--verbose` | Enable verbose mode with detailed logs including prompts. |
| `--disable-chat-ui` | Disable the Chat UI served at the root path. |
| `--auto-reload` | Enable auto reload when configuration files change. |
| `--prefix` | A prefix to add to all server paths. Must start with `/`. |
***
## `eval`
Evaluate guardrail configurations with various testing and compliance workflows.
```bash
nemoguardrails eval run --config examples/configs/content_safety
```
### Subcommands
| Subcommand | Description |
| ------------------ | ------------------------------------------------------------------- |
| `run` | Run the interactions for an evaluation. |
| `check-compliance` | Check the policy compliance of the interactions in the output path. |
| `ui` | Launch the Evaluation UI. |
| `rail` | Run a rail evaluation task. |
Use `nemoguardrails eval --help` for details on each subcommand.
***
## `convert`
Convert Colang files and configurations from older versions to the latest format.
```bash
nemoguardrails convert ./my-config --from-version 1.0 --validate
```
### Arguments
| Argument | Description |
| -------- | ------------------------------------------------------- |
| `PATH` | The path to the file or directory to migrate. Required. |
### Options
| Option | Description |
| ------------------------ | --------------------------------------------------------------------------------------------- |
| `--from-version` | The version of the Colang files to migrate from. Options: `1.0`, `2.0-alpha`. Default: `1.0`. |
| `--verbose` | Enable verbose mode with detailed logs. |
| `--validate` | Validate the output using the Colang Parser. |
| `--use-active-decorator` | Use the active decorator in the migration. Default: enabled. |
| `--include-main-flow` | Add a main flow to the config. Default: enabled. |
***
## `find-providers`
List and select LLM providers interactively. This command helps you discover available providers for text completion and chat completion models.
```bash
nemoguardrails find-providers --list
```
### Options
| Option | Description |
| -------------- | ----------------------------------------------------------- |
| `--list`, `-l` | List all available providers without interactive selection. |
### List Mode
Run the following command to list all available providers:
```bash
nemoguardrails find-providers --list
```
### Interactive Mode
Run the command without options to start an interactive provider selection:
```bash
nemoguardrails find-providers
```
1. Select a provider type:
* Type to filter between "text completion" and "chat completion".
* Use arrow keys to navigate through matches.
* Press Tab to autocomplete.
* Press Enter to select.
2. Select a specific provider:
* Type to filter through available providers.
* Use arrow keys to navigate through matches.
* Press Tab to autocomplete.
* Press Enter to select.
#### Example
```text
Available Provider Types: (type to filter, use arrows to select)
• text completion
• chat completion
Select provider type: text
Available text completion providers: (type to filter, use arrows to select)
• anthropic
• google_palm
• openai
• ...
Select provider: openai
Selected text completion provider: openai
```