svelte-debugging

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill documents defensive programming practices and security best practices for Svelte development. It specifically advises against exposing secrets in client-side bundles, implementing proper CORS and CSRF protections, and using secure cookie attributes (HttpOnly, Secure, SameSite).
  • [INDIRECT_PROMPT_INJECTION]: The skill's primary function involves analyzing untrusted inputs such as application source code, execution logs, and API traces to identify bugs.
  • Ingestion points: The 'Process' section in SKILL.md instructs the agent to gather evidence from code, logs, traces, metrics, and generated output.
  • Boundary markers: The skill relies on the agent's native handling of these inputs and does not define specific delimiters for log data.
  • Capability inventory: The skill provides instructions for manual analysis and code remediation; it does not include scripts for automated command execution or network exfiltration.
  • Sanitization: The reference documentation (references/01-overview.md, references/02-autoryzacja-i-sesja.md) strongly advocates for the use of validation libraries like Zod and Valibot to sanitize data at application boundaries.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 02:31 PM
Security Audit — agent-trust-hub — svelte-debugging