writing-typescript-code
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill documents standard project maintenance commands, including
npm run devfor starting the Vite development server andnpm cifor performing clean dependency installations based on the lock file. - [INDIRECT_PROMPT_INJECTION]: The architecture described for components like
Markdown.tsxandAssistantMessage.tsxinvolves processing and rendering content retrieved from external sources such as the Azure AI Agent API. - Ingestion points: The UI ingests and renders content from external chat and response APIs through the
Markdown.tsxcomponent (SKILL.md). - Boundary markers: The skill does not explicitly define specific delimiters or boundary markers for untrusted content within the UI instructions.
- Capability inventory: The frontend architecture performs authenticated network requests via the
fetchAPI and handles token acquisition through MSAL (SKILL.md). - Sanitization: The guidelines focus on type safety and component structure, but do not specify explicit sanitization or escaping procedures for the rendering of chat content.
Audit Metadata