web-meta-framework-astro
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill documents methods for ingesting data from external APIs and rendering it directly into components (e.g., in
examples/content.mdandexamples/routing.md). This design creates a vulnerability surface where instructions embedded in external data could be processed as commands by an AI agent if boundaries or sanitization are not applied. - Ingestion points: Multiple examples demonstrate fetching remote content using
fetch()in content loaders, SSR pages, and API endpoints (e.g.,https://api.example.com/products). - Boundary markers: The provided code patterns do not include delimiters or instructions for the agent to ignore embedded instructions within the fetched data.
- Capability inventory: The skill enables server-side execution, full-stack routing, and the ability to interact with various web services.
- Sanitization: The examples show direct interpolation of external data into HTML templates without explicit sanitization or validation steps for security.
Audit Metadata