sveltekit-remote-functions
Pass
Audited by Gen Agent Trust Hub on Mar 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill acts as a technical reference for SvelteKit development, promoting security-first patterns such as schema-based input validation.
- [PROMPT_INJECTION]: The skill describes an architecture for processing untrusted data with potential injection surfaces. 1. Ingestion points: Remote function arguments (e.g., in delete_user, create_post, get_weather) and database content rendered via @html post.content. 2. Boundary markers: The skill uses standard markdown code fences and specifies file naming conventions (*.remote.ts) to isolate logic. 3. Capability inventory: Includes database access (db.users.delete), cookie access (getRequestEvent().cookies.get), and server-side state modification. 4. Sanitization: Mandates the use of StandardSchemaV1 (Valibot/Zod) for all function inputs to validate and sanitize data before processing.
- [DATA_EXFILTRATION]: No malicious exfiltration patterns were detected. The skill describes legitimate cookie and header access for server-side authorization and session management.
- [CREDENTIALS_UNSAFE]: No hardcoded secrets were found. The documentation specifically instructs developers to use private field naming (e.g., _password) to prevent sensitive data from being returned to the client.
Audit Metadata