1# Azure AI Content Safety — TypeScript SDK Quick Reference
2 
3> Condensed from **azure-ai-contentsafety-ts**. Full patterns (blocklist CRUD, image moderation, severity thresholds)
4> in the **azure-ai-contentsafety-ts** plugin skill if installed.
5 
6## Install
7```bash
8npm install @azure-rest/ai-content-safety @azure/identity @azure/core-auth
9```
10 
11## Quick Start
12```typescript
13import ContentSafetyClient, { isUnexpected } from "@azure-rest/ai-content-safety";
14import { AzureKeyCredential } from "@azure/core-auth";
15const client = ContentSafetyClient(endpoint, new AzureKeyCredential(key));
16```
17 
18## Non-Obvious Patterns
19- REST client — `ContentSafetyClient` is a function, not a class
20- Text: `client.path("/text:analyze").post({ body: { text, categories: [...] } })`
21- Image: `client.path("/image:analyze").post({ body: { image: { content: base64 } } })`
22- Blocklist create: `.path("/text/blocklists/{blocklistName}", name).patch({...})`
23- API key import: `AzureKeyCredential` from `@azure/core-auth` (not `@azure/identity`)
24 
25## Best Practices
261. Always use `isUnexpected()` — type guard for error handling
272. Set appropriate thresholds — different categories may need different severity levels
283. Use blocklists for domain-specific terms to supplement AI detection
294. Log moderation decisions — keep audit trail for compliance
305. Handle edge cases — empty text, very long text, unsupported image formats
31