gws-modelarmor
Google Model Armor: Filter user-generated content for safety.
57
Quality
47%
Does it follow best practices?
Impact
68%
1.47xAverage score across 3 eval scenarios
Optimize this skill with Tessl
npx tessl skill review --optimize ./skills/gws-modelarmor/SKILL.mdEvaluation results
75%
25%Model Armor Template Audit Script
Schema-first API discovery
Help discovery step
100%
100%
Schema inspection step
0%
100%
Schema-driven flags
0%
50%
gws-shared prerequisite
33%
0%
Uses gws CLI commands
100%
100%
JSON output produced
100%
100%
Without context: $0.2270 · 1m 3s · 15 turns · 21 in / 3,368 out tokens
With context: $0.3898 · 1m 20s · 24 turns · 29 in / 4,278 out tokens
65%
10%LLM Content Safety Integration Script
Prompt and response sanitization helpers
Sanitize-prompt for user input
40%
60%
Sanitize-response for LLM output
40%
60%
Both stages present
100%
100%
Schema discovery used
0%
0%
Template referenced
100%
100%
Does NOT use raw REST/HTTP
100%
100%
Without context: $0.3755 · 1m 39s · 19 turns · 25 in / 5,894 out tokens
With context: $0.5536 · 2m 9s · 34 turns · 338 in / 7,474 out tokens
65%
32%Model Armor Safety Template Setup
Template creation with helper command
Uses create-template helper
0%
0%
Help discovery step
100%
100%
Schema inspection step
0%
100%
Schema-driven flags
0%
100%
gws-shared prerequisite
30%
0%
Does NOT use direct REST API
100%
100%
Without context: $0.4692 · 2m 26s · 34 turns · 175 in / 7,921 out tokens
With context: $0.6336 · 2m 45s · 37 turns · 37 in / 7,938 out tokens
- Evaluated
- Agent
- Claude Code
- Model
- Claude Sonnet 4.6
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.