PluginBench
Skill
Pass
Audit score 90

gws-modelarmor

googleworkspace/cli

Filter user-generated content for safety using Google Model Armor templates.

What is gws-modelarmor?

Google Model Armor is a safety filtering service that sanitizes user prompts and model responses through configurable templates. Use it to protect applications from harmful content and ensure compliance with safety policies.

  • Sanitize user prompts before sending to models
  • Filter model responses for safety violations
  • Create and manage custom Model Armor templates
  • Inspect API schemas to understand required parameters
  • Apply safety policies consistently across interactions

How to install gws-modelarmor

npx skills add null --skill gws-modelarmor
Prerequisites
  • Google Workspace CLI (gws) installed and configured
  • Authentication set up via gws-shared skill
  • Access to Google Model Armor API
Claude Code
Cursor
Windsurf
Cline

How to use gws-modelarmor

  1. 1.Run `gws modelarmor --help` to browse available resources and methods
  2. 2.Use `gws schema modelarmor.<resource>.<method>` to inspect required parameters
  3. 3.Call `gws modelarmor sanitize-prompt` to filter user input before processing
  4. 4.Call `gws modelarmor sanitize-response` to filter model output
  5. 5.Create custom templates with `gws modelarmor create-template` for organization-specific policies

Use cases

Good for
  • Sanitize user input in chatbot applications before processing
  • Filter AI-generated responses to remove harmful content
  • Create organization-specific safety templates for content moderation
  • Validate prompts and responses against custom safety policies
  • Protect user-facing applications from unsafe content generation
Who it's for
  • AI application developers
  • Content moderation teams
  • Security and compliance engineers
  • Teams building chatbots or generative AI products

gws-modelarmor FAQ

What is the difference between sanitize-prompt and sanitize-response?

sanitize-prompt filters user input before it reaches the model, while sanitize-response filters the model's output before returning it to users.

Can I create custom safety templates?

Yes, use the create-template command to build organization-specific Model Armor templates tailored to your safety requirements.

Do I need additional authentication beyond gws-shared?

No, authentication is handled by the gws-shared skill. Ensure it is installed and configured first.

How do I discover available API methods?

Run `gws modelarmor --help` to browse resources, then use `gws schema modelarmor.<resource>.<method>` to inspect parameters.

Full instructions (SKILL.md)

Source of truth, from googleworkspace/cli.


name: gws-modelarmor description: "Google Model Armor: Filter user-generated content for safety." metadata: version: 0.22.5 openclaw: category: "productivity" requires: bins: - gws cliHelp: "gws modelarmor --help"

modelarmor (v1)

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

gws modelarmor <resource> <method> [flags]

Helper Commands

CommandDescription
+sanitize-promptSanitize a user prompt through a Model Armor template
+sanitize-responseSanitize a model response through a Model Armor template
+create-templateCreate a new Model Armor template

Discovering Commands

Before calling any API method, inspect it:

# Browse resources and methods
gws modelarmor --help

# Inspect a method's required params, types, and defaults
gws schema modelarmor.<resource>.<method>

Use gws schema output to build your --params and --json flags.

gws-modelarmor — AI Skill | PluginBench