untrusted-content

Here are 2 public repositories matching this topic...

Semantic-layer prompt injection defence that separates untrusted instructions from authority while preserving the legitimate task.

Security-focused JSON UI DSL, validator, schemas, and React renderer for untrusted user- and LLM-generated content.

react security typescript json-schema declarative-ui ugc llm generative-ui safe-rendering untrusted-content

To associate your repository with the untrusted-content topic, visit your repo's landing page and select "manage topics."