1---
2name: gws-modelarmor-sanitize-response
3description: "Google Model Armor: Sanitize a model response through a Model Armor template."
4metadata:
5  version: 0.22.5
6  openclaw:
7    category: "security"
8    requires:
9      bins:
10        - gws
11    cliHelp: "gws modelarmor +sanitize-response --help"
12---
13 
14# modelarmor +sanitize-response
15 
16> **PREREQUISITE:** Read `../gws-shared/SKILL.md` for auth, global flags, and security rules. If missing, run `gws generate-skills` to create it.
17 
18Sanitize a model response through a Model Armor template
19 
20## Usage
21 
22```bash
23gws modelarmor +sanitize-response --template <NAME>
24```
25 
26## Flags
27 
28| Flag | Required | Default | Description |
29|------|----------|---------|-------------|
30| `--template` | ✓ | — | Full template resource name (projects/PROJECT/locations/LOCATION/templates/TEMPLATE) |
31| `--text` | — | — | Text content to sanitize |
32| `--json` | — | — | Full JSON request body (overrides --text) |
33 
34## Examples
35 
36```bash
37gws modelarmor +sanitize-response --template projects/P/locations/L/templates/T --text 'model output'
38model_cmd | gws modelarmor +sanitize-response --template ...
39```
40 
41## Tips
42 
43- Use for outbound safety (model -> user).
44- For inbound safety (user -> model), use +sanitize-prompt.
45 
46## See Also
47 
48- [gws-shared](../gws-shared/SKILL.md) — Global flags and auth
49- [gws-modelarmor](../gws-modelarmor/SKILL.md) — All filter user-generated content for safety commands
50