better-auth

Pass

Audited by Gen Agent Trust Hub on Jun 20, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Recommends the installation of better-auth-drizzle-adapter, a community-maintained package by the author (remorses). The skill provides technical justification for this recommendation, citing specific bug fixes for Drizzle v1 and Postgres-js not yet present in the official adapter. The package source is transparently linked to GitHub.
  • [COMMAND_EXECUTION]: Suggests using pnpm dlx auth@latest generate to automate database schema creation. This is a standard utility command provided by the better-auth library for development workflows.
  • [CREDENTIALS_UNSAFE]: Provides guidance on managing sensitive environment variables like BETTER_AUTH_SECRET and GOOGLE_CLIENT_SECRET. The instructions correctly emphasize that these must be treated as secrets, managed via secure tools (such as Doppler or Sigillo), and never hardcoded.
  • [DATA_EXFILTRATION]: Includes instructions for forwarding headers and cookies during server-side API calls. This is a legitimate functional requirement for maintaining authentication state and CSRF protection during OAuth flows, rather than an exfiltration risk.
  • [INDIRECT_PROMPT_INJECTION]: The skill describes authentication flows that ingest data from external social providers (Google, GitHub) and user registration forms.
  • Ingestion points: User-supplied names and emails during sign-up; profile data returned from OAuth providers (Google, GitHub).
  • Boundary markers: Demonstrates the use of Zod schemas to validate and delimit incoming request data in route handlers.
  • Capability inventory: Performs database operations (Drizzle ORM) and handles HTTP redirects.
  • Sanitization: Input validation is consistently implemented using Zod schema verification across the provided code examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 20, 2026, 06:48 PM
Security Audit — agent-trust-hub — better-auth