# Gws Modelarmor Sanitize Prompt

> Google Model Armor: Sanitize a user prompt through a Model Armor template.

Source: https://skillsagentes.com/skills/googleworkspace/cli/gws-modelarmor-sanitize-prompt
Repository: https://github.com/googleworkspace/cli
Author: googleworkspace
License: Apache-2.0
Updated: hace 4 meses
Context cost: 19 tok installed, 358 tok once triggered, 358 tok with every bundled file
Bundle: 1 file, 1 KB
Permissions requested: none declared

## Install

```bash
npx -y skills add googleworkspace/cli --skill gws-modelarmor-sanitize-prompt --agent claude-code
```

## What it does

- Ejecuta gws modelarmor +sanitize-prompt para pasar un prompt de usuario por una plantilla de Model Armor de Google Cloud
- Acepta el texto vía --text, --json o desde stdin si no se indica ninguno

## Use it when

- Se necesita sanitizar el prompt de un usuario antes de enviarlo a un modelo
- Se quiere filtrar contenido de entrada usando una plantilla de Model Armor existente

## Don't bother when

- Para la seguridad de la salida del modelo, en cuyo caso se debe usar +sanitize-response en su lugar

## What triggers it

- "Sanitiza este prompt de usuario con la plantilla projects/P/locations/L/templates/T"
- "Pasa este texto por Model Armor antes de enviarlo al modelo"

## Before you install

- Requiere el binario gws y haber leído ../gws-shared/SKILL.md para autenticación, flags globales y reglas de seguridad.

## Files

- SKILL.md — 1 KB

## SKILL.md

Reproduced verbatim from googleworkspace/cli under Apache-2.0. This section is the upstream document and is in English.

# modelarmor +sanitize-prompt

> **PREREQUISITE:** Read `../gws-shared/SKILL.md` for auth, global flags, and security rules. If missing, run `gws generate-skills` to create it.

Sanitize a user prompt through a Model Armor template

## Usage

```bash
gws modelarmor +sanitize-prompt --template <NAME>
```

## Flags

| Flag | Required | Default | Description |
|------|----------|---------|-------------|
| `--template` | ✓ | — | Full template resource name (projects/PROJECT/locations/LOCATION/templates/TEMPLATE) |
| `--text` | — | — | Text content to sanitize |
| `--json` | — | — | Full JSON request body (overrides --text) |

## Examples

```bash
gws modelarmor +sanitize-prompt --template projects/P/locations/L/templates/T --text 'user input'
echo 'prompt' | gws modelarmor +sanitize-prompt --template ...
```

## Tips

- If neither --text nor --json is given, reads from stdin.
- For outbound safety, use +sanitize-response instead.

## See Also

- [gws-shared](../gws-shared/SKILL.md) — Global flags and auth
- [gws-modelarmor](../gws-modelarmor/SKILL.md) — All filter user-generated content for safety commands

---

Skills Agentes — https://skillsagentes.com/skills/googleworkspace/cli/gws-modelarmor-sanitize-prompt