astro
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill documents patterns for ingesting and processing untrusted data via framework-specific request handling mechanisms.
- Ingestion points: Request headers and cookies are accessed in
references/middleware.md, and POST body and form data are processed inreferences/astro-db.md. - Boundary markers: The provided code snippets do not contain explicit delimiters or instructions for the agent to treat interpolated content as untrusted.
- Capability inventory: The skill demonstrates capabilities for database writes using
astro:db, programmatic navigation/redirects, and response header modification. - Sanitization: Example code focused on framework features does not demonstrate sanitization or validation of untrusted input before persistence or usage.
- [EXTERNAL_DOWNLOADS]: The documentation provides instructions for setting up standard framework integrations and libraries from well-known registries.
- Evidence: Includes commands such as
npx astro addfor official adapters and integrations (React, Svelte, Vue, Tailwind, MDX, DB) andnpm installfor community state management tools like Nanostores.
Audit Metadata