azure-ai-contentsafety-py

Azure AI Content Safety SDK for Python. Use for detecting harmful content in text and images with multi-severity classification.

Install with Tessl CLI

npx tessl i github:boisenoise/skills-collections --skill azure-ai-contentsafety-py

What are skills?

1.07x

Quality

68%

Does it follow best practices?

Impact

100%

1.07x

Average score across 3 eval scenarios

Optimize this skill with Tessl

npx tessl skill review --optimize ./skills/antigravity-azure-ai-contentsafety-py/SKILL.md

SKILL.md

Review

Evals

Evaluation results

100%

Community Forum Moderation Tool

Text content analysis with category severity

Criteria

Without context

With context

Correct package import

100%

ContentSafetyClient used

100%

AzureKeyCredential auth

100%

CONTENT_SAFETY_ENDPOINT env var

100%

CONTENT_SAFETY_KEY env var

100%

AnalyzeTextOptions model

100%

analyze_text method

100%

All four TextCategories

66%

100%

categories_analysis response field

100%

Severity reported per category

100%

Multiple categories handled

100%

Without context: $0.2297 · 3m 57s · 19 turns · 26 in / 2,993 out tokens

With context: $0.4041 · 3m 45s · 24 turns · 130 in / 4,376 out tokens

100%

E-commerce Platform Custom Content Filtering

Custom blocklist creation and moderation

Criteria

Without context

With context

BlocklistClient import

100%

BlocklistClient instantiation

100%

TextBlocklist model used

100%

create_or_update_text_blocklist called

100%

AddOrUpdateTextBlocklistItemsOptions used

100%

TextBlocklistItem used

100%

add_or_update_blocklist_items called

100%

blocklist_names in AnalyzeTextOptions

100%

halt_on_blocklist_hit set

100%

blocklists_match checked

100%

blocklist_item_text accessed

100%

Env vars used

100%

Without context: $0.5154 · 6m 7s · 30 turns · 33 in / 6,825 out tokens

With context: $0.3763 · 3m 26s · 21 turns · 447 in / 4,123 out tokens

100%

16%

AI Writing Assistant Response Filter

AI output pre-moderation with fine-grained severity

Criteria

Without context

With context

EIGHT_SEVERITY_LEVELS output type

66%

100%

AnalyzeTextOutputType imported

100%

Pre-moderation before display

100%

Severity threshold applied

100%

All four TextCategories checked

100%

ContentSafetyClient used

100%

AnalyzeTextOptions model

100%

Env vars for credentials

50%

100%

Analysis results logged

100%

AzureKeyCredential used

100%

categories_analysis accessed

100%

Without context: $0.2453 · 2m 18s · 17 turns · 20 in / 3,499 out tokens

With context: $0.3822 · 3m 24s · 21 turns · 24 in / 4,301 out tokens

Evaluated: 4 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Community Forum Moderation Tool E-commerce Platform Custom Content Filtering AI Writing Assistant Response Filter

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.